right? 04:14:14 then after the HTML onload, JS asks for the data and fills in the dom? 04:14:43 regretfully I haven't tested the firefox in a while 04:16:05 melvster and I focused on getting editable working again via SPARQL this week. glad to have you back! 04:16:14 Unfortunately Firefox doesn't gove you a stack dump 04:16:43 I see it referenced already in tabulator.xul 04:17:05 I thought this was the server/component "window" 04:17:25 it? jQuery? 04:17:35 yes line 9 of tabulator.xul overlay 04:17:47 The firefox component ius in components/xpcom.js 04:19:59 maybe the current script tag can be moved there then 04:21:08 I've never looked closely at the tabulator.xul, I've only tweaked it to add a menu item 04:21:26 If it is menus it must be a level above any one tab or window 04:21:48 theres also an HTML doc inline in getOutlinerHTML 04:22:03 of xpcom.js, line 48 04:22:53 Yes. That outliner HTML is put in every single tabulator window (or tab) 04:23:07 possible the jquery needs to get loaded either there or js/init/init.js 04:24:04 I thought Jim added it there ... I wonder whether getOutlinerHTML is used 04:27:11 lkagal has quit (Quit: lkagal) 04:27:29 / Now we have to load jQuery for rda -- but it is also loaded in ../../tabulator.xul @@ 04:27:29 / tabulator.loadScript("js/jquery/jquery-1.4.2.min.js"); 04:27:42 in init.js 04:28:23 hm then we should already have the necessary bits everywhere in all cases? 04:28:50 also, http://blog.nodejitsu.com/jsdom-jquery-in-5-lines-on-nodejs 04:29:47 though I'm still not sure why dom is needed for XHR 04:30:23 It isn't -- it is needed for jQuery and ... hence not putting a dependency on jQuery into a basic RDF ibrary 04:30:48 Using jQuery for XHR brings in a lot of stuff, such as the DOM 04:31:23 Idea was you could use turtle without having to implement the dom if you don't need RDF/a or RDF/xml 04:31:41 good idea, sorry feel free to revert :) 04:32:17 the proxy thing was nice though... haven't had to fight anyone for CORS in months! 04:32:23 jQuery needs to be refactored IMHO if the XHR stuff is so useful 04:33:04 Maybe we could grab the XHR bits outof jQuery without the DOM dependency 04:33:18 proxy eg: http://data.fm/proxy?uri=http://www.w3.org/People/Berners-Lee/card 04:35:05 it has a nice plugin interface, data filters, and more 04:36:01 WHat should I get with http://data.fm/proxy?uri=http://www.w3.org/People/Berners-Lee/card ? 04:36:10 tabulator should load 04:36:49 Tabulator does but there is no data displayed 04:37:01 The request seems to fail 04:37:24 Request for http://data.fm/proxy?uri=http://www.w3.org/People/Berners-Lee/card no status 04:37:30 try chrome? 04:37:39 That was on firefox 04:37:52 Safari asks for a private key without explaining why 04:38:10 probably your foaf links to HTTPS 04:38:52 Application: WebProcess Keychain: timbl.keychain does not tell me enough 04:39:22 info about the XHR would be nice 04:39:52 What web site is it going to use the key on, yes 04:40:08 which key did it want to use? 04:40:28 ok, i f I tellit allow I get a tabulator output on safari 04:40:29 which method is being called would be nice too 04:40:39 yes 04:40:55 and by code from what domain etc 04:42:09 Chrome a bit better 04:43:25 also now there's an editable indicator on the internals pane 04:43:38 Maybe the thing to do in web.js is to make it conditional if jQuery is defined 04:43:58 use it, else use the XMLXHR factory in the library 04:44:31 editable indicatro: nice 04:45:38 Why does it say that http://data.fm/proxy?uri=http://www.w3.org/People/Berners-Lee/card is editable? 04:46:31 the HTTP header says DAV, SPARQ 04:46:36 the server advertised that errantly 04:46:47 but it is a proxy to my card which should not be eitable 04:46:58 error in the proxy 04:47:32 you should also see tabulator here: http://tabulator.data.fm/graph 04:47:35 shouldn't the internal pane have said "DAV, SPARQ" 04:48:07 that one really is editable 04:48:56 Yes I see /graph 04:49:21 the call editable(graph) only returns its first preference for editing, not all available 04:49:54 ok 04:50:06 of course the whole editing thing needs overhaul too 04:51:58 because of the "don't know yet problem? 04:52:46 yes also the events bind to getEltById(docHTML) 04:53:01 so its not very portable and editing only works if your skin is just right 04:53:15 yes 04:53:38 in fact we should generalize it so you can make all kidns of mashups 04:54:00 not just one in the page, and the tabulator default view is just a trvial case 04:55:46 BTW Is there a way of going a git pull origin master recursively in submodules ? 04:56:17 js/tab/outline.js#L2132 04:56:36 silently fails editing via kb/mouse 04:57:22 re git. I think tabulator has a make target for that 04:58:19 outline.js had a bunch of catch(e) {} which we should erradicate 04:59:23 because the default error handling on firefox doesn't give yo the stack, it maybe best to replace them with a specific output of the stack 04:59:45 the code which calls the pain uses a utility routine to do that 04:59:53 IIRC 05:00:00 github (~github@sh2.rs.github.com) has joined #dig 05:00:00 [tabulator] presbrey pushed 1 new commit to master: http://git.io/vhUuXw 05:00:00 [tabulator/master] Makefile: drop old all target - presbrey 05:00:00 github has left #dig 05:01:32 it seems >10% is old code commented out, should go too 05:02:12 Well, unless it was code which is needd and soemone had not got working yet 05:02:24 re git 05:02:36 git submodule update --init 05:02:41 is not recursive, is it? 05:03:09 it just reads the current project .gitmodules file 05:03:23 so not submodules within submodules 05:05:01 I see so so you have knitted it together thorugh Makefiles trough the tree 05:05:12 thank you! Pity git couldn't do that 05:06:39 yes its submodule support seems lackluster 05:09:07 bolted on afterwards and not really thought out from the POV of an actual user. 05:11:07 It is an art to make a ui which makes simplest the things people actually want to do easy. 05:11:47 I think jquery is nice to that end in that it stays out of the way 05:12:06 eg. tabulator does not (nor need to) know it has proxy support 05:12:17 yes 05:12:21 latest mashlib skin: https://github.com/linkeddata/data.fm/blob/master/www/inc/contrib/skin.html.php#L16 05:12:47 if you see the eg. actually jquery doesn't really know about the proxy either 05:13:48 I'm going to head off now, tomorrow I need to use the tabulator for my taxes so I'll look at making jquery be optional or something 05:14:34 great I'll be around to hack too, good night 05:16:08 great 05:23:59 scor has quit (Quit: scor) 07:14:08 melvster (~melvin@p5797EBD2.dip.t-dialin.net) has joined #dig 10:56:35 danbri has quit (Remote host closed the connection) 10:58:04 danbri (~danbri@cable-146-255-149-19.dynamic.telemach.ba) has joined #dig 11:54:17 danbri has quit (Read error: Connection reset by peer) 12:02:42 danbri (~danbri@cable-146-255-149-19.dynamic.telemach.ba) has joined #dig 13:56:38 tlr (~tlr@88.207.137.200) has joined #dig 16:08:45 lkagal (~lkagal@pool-96-237-240-136.bstnma.fios.verizon.net) has joined #dig 16:54:01 lkagal has quit (Quit: lkagal) 16:57:49 git still driving me crazy 17:42:35 var req = kb.bnode(); // @@ Joe, no need for xhr.req? 17:45:09 danbri has quit (Remote host closed the connection) 17:45:17 don't see that comment in my branch 17:46:07 ./js/rdf/web.js:607: var req = xhr.req = kb.bnode() 17:49:39 oh I was behind... 17:49:46 ./web.js:612: var req = kb.bnode() 17:49:53 there is no need there because xhr isn't created yet 17:50:21 it starts here: 17:50:26 925 var xhr = jQuery.ajax({ 17:52:17 there are a couple factories I created 17:52:24 that abstract onerror/onreadystate from the xhr 17:52:37 this should be very useful in making jquery optional 17:52:58 infact we can probably do it in very few lines ~925 17:55:12 prior to Factory-style, the event methods were set directly on xhr which ties too tightly to use multiple xhr implementations 17:55:41 eg. since jquery does not handle 'xhr.onerror =' setups 18:01:49 Ok, I have left int the factories 18:01:56 left in 18:03:11 I have in the non-jquery case set up xhr.uri and xhr.req early on in case some version of XHR calls an error handler imemdiaely -- the erro handler will need xhr.req defined 18:03:33 danbri (~danbri@cable-146-255-149-19.dynamic.telemach.ba) has joined #dig 18:04:19 do you know which errors? 18:04:40 it would be nice to setup a non-jquery/node.js environ to simulate these cases 18:05:36 We should have a node-based test suite for the whole lib 18:07:43 for rdflib.js, yes 18:07:43 If you were to implement jq.ajaz in a system with no asynch callback, presumably you'd have to call the callbacks before jq.ajax returned 18:08:08 which JS system is that with no async callback? 18:09:03 None that I know of but who knows 18:09:17 in tabulator firefox 10.0.1 I am getting the download box on auth/acl.n3 18:09:33 I think that is not an rdflib.js issue btw 18:09:42 (not jquery either) 18:10:08 I have a patched version now which only uses jQuery is it is defined 18:10:43 I do not get the file download box. I get tabulator, but the http request never seems to get its repsonse 18:10:49 o i have no data 18:11:17 Must have missed something 18:12:08 its fine to have jquery be optional anyway 18:12:29 so you should commit that soon so we don't have to merge later :) 18:12:38 I think it is good to keep deependencies down if we want it picked up widely 18:12:57 commit before it works? 18:13:10 ha then you can help find the problem 18:13:22 first I have to fix my remote so it can push 18:13:22 no sorry, new branch if you want to do that 18:13:34 after you test it if you should send to master 18:13:56 Suppose I want t push using an SSH key? 18:14:30 did you checkout a read-only ? 18:14:44 edit your remote origin if so 18:15:28 I checked out from a read-write module using git sumbodule -- so the submodule system is set up to be read-only I giuess 18:17:29 you probably want to edit content/js/rdf/.git/config 18:18:15 [origin] url = git@github.com:linkeddata/rdflib.js.git 18:18:54 git remote set-url --push origin https://github.com/linkeddata/rdflib.js.git 18:19:16 $ git remote -v 18:19:16 origin git://github.com/linkeddata/rdflib.js.git (fetch) 18:19:16 origin https://github.com/linkeddata/rdflib.js.git (push) 18:19:45 nice but https:// doesn't use your keys 18:20:52 no, if I use ssh://timbl@github.com/ can I still push to ;linkeddata ? 18:21:30 maybe ssh://timbl@github.com/linkeddata/rdflib.js.git 18:22:15 you have to use git@github.com:linkeddata/rdflib.js.git 18:22:42 also it looks like menu items have changed in firefox 10.0.1: 18:22:54 Error: undefined entity 18:22:54 Source File: chrome://tabulator/content/tabulator.xul 18:22:54 Line: 59, Column: 9 18:22:54 Source Code: 18:22:54 18:22:57 why git@ ? 18:23:22 Yes, there are two lines missing from the DTD 18:23:41 I have fixed them and will cehck them in -- dunno why the vanished, they were there 18:23:54 why git@ instead iof timbl@ ? 18:24:07 my keys are under timbl surely 18:24:23 ssh://timbl@github.com/linkeddata/rdflib.js.git won't work? 18:24:23 there is no timbl@github.com afaik 18:24:49 try with the plain ssh client, ssh timbl@github.com. vs. ssh git@gitub.com. 18:32:44 and that should use my SSH keys? 18:34:28 ok, I pushd the dtd to git@github.com:linkeddata/tabulator-firefox.git 18:48:59 ok now its working, loading data, and I get: 18:49:02 @@ xpcom.js test 92 $rdf.log.error)function (msg) { 18:49:02 this.msg(msg, this.TERROR, "eror", arguments); 18:49:02 } 18:49:02 @@ xpcom.js test 94 $rdf.log.error)function (msg) { 18:49:03 this.msg(msg, this.TERROR, "eror", arguments); 18:49:05 } 18:49:24 I do see data in firefox 10.0.1 18:49:35 I get the HTTP Request Sent status line, so the original call runs to completio 18:49:46 This is the extension r the mashup? 18:49:50 working for you 18:50:14 extension @ http://www.w3.org/ns/acl/auth.n3 18:50:47 You make jQuery optional? 18:51:23 made 18:51:38 no changes, all latest code 18:52:20 tabulator-firefox]$ git branch -v; cd content/; git branch -v; cd js/rdf; git branch -v 18:52:20 * master 6f78427 Ooops misssing define tabulatorSaveSession.* entities 18:52:20 * master dd00a91 Makefile: drop old all target 18:52:20 * master aa28203 Change INSERT DATA to INSERT INTO <> 18:52:25 Guess I have to upgrade from 8 to 10 18:54:58 or 11? 18:55:27 :/ I have 10.0.1 19:02:32 how do I revert to the same that you have then -- git revert origin master ? 19:06:31 check what branch you are on vs. available with: git branch -av 19:06:38 you have to check for each submodule too 19:06:52 if you are not on master, do: git checkout master 19:07:06 to sync: git pull origin master 19:07:16 to throw away your changes: git reset --hard origin/master 19:07:42 reset is local only, so make sure you have latest master locally first or you will be reseting to old stuff 19:09:41 I have 10.0.2 and 11.0b5 19:12:19 Already up-to-date. 19:12:19 $ git reset --hard origin/master 19:12:19 HEAD is now at aa28203 Change INSERT DATA to INSERT INTO <> 19:12:58 timbl: that's a change we made lately ... we're starting to get some edit functionality going on data.fm 19:13:15 ps welcome back! :) 19:13:54 Is INSERT DATA no longer part of SPRQL? 19:14:01 thank you! 19:14:04 right now I can add a triple on data.fm using mashlib.js using the magnifying glass and the big blue plus ... I'm looking to see how we can get working again to change a triple 19:14:34 we had to use some parts of ARC2 which has the old sparql plus syntax 19:14:52 Ok, so now I use FF10 and the current origin master I still get no data if I run /Applications/Firefox10.app/Contents/MacOS/firefox-bin -jsconsole http://www.w3.org/People/Berners-Lee/card.n3 19:15:07 kingsley said that INSERT INTO <> is valid sparql that will work with both sparql update and ARC2 ... sort of a hack but it's starting to work... 19:16:20 i havent tested the extension on ff10 yet ... been mainly working with mashlib and chrome, as I figure that's a nice thing to demo to people that are less familiar with tabulator 19:17:19 I needed to do my taxes this weekend and I can't do it until I get tabulator-firefox going again 19:17:34 ah! 19:17:35 so I have a specific motivation 19:17:55 i noticed the financial tab ... nice job :) 19:18:10 pane i mean 19:18:51 I've been doing finances in RDF for ages I should probably move the OFX parser into tabulator too 19:19:16 now it is a mix of tabulator and cwm and a pyhtion ofx to rdf converter 19:19:25 nice dogfooding! :) 19:20:53 scor (~scor@drupal.org/user/52142/view) has joined #dig 19:22:58 So Prebrey, you say that with FFox 10, and the current latest code it works? Sure no diffs? 19:25:00 github (~github@sh2.rs.github.com) has joined #dig 19:25:00 [tabulator-firefox] timbl pushed 1 new commit to master: http://git.io/jOqPtw 19:25:00 [tabulator-firefox/master] Say it supports Firefox 11 - Tim Berners-Lee 19:25:00 github has left #dig 19:28:29 git continues to drive me crazy 19:29:01 $ git status 19:29:01 # Not currently on any branch. 19:29:01 # Changes not staged for commit: 19:29:02 # 19:29:02 # modified: js/rdf (new commits) 19:29:02 # 19:29:03 no changes added to commit (use "git add" and/or "git commit -a") 19:29:03 $ cd js/rdf 19:29:03 $ git status 19:29:03 # Not currently on any branch. 19:29:04 nothing to commit (working directory clean) 19:29:04 $ 19:29:06 for example 19:29:33 It says the submodule has comits but you go there and it hasn't 19:33:43 one possibility is to clone a fresh copy (and still keep the old directory) ... git clone --recursive git@github.com:linkeddata/tabulator-firefox.git 19:36:17 --recursive will pull in the submodules? 19:36:23 yes it should do 19:42:35 Ok, phantom "content modified" has gone 19:42:58 no way to run a restaurant having to re-clone like that to get it consistent 19:43:02 git-- 19:50:33 git is quite clever it has 3 types of objects (1. sha1 of a file, 2. sha1 of a file system tree which points to the files, 3. sha1 of a commit which adds the check in information) ... branches are just a short form for a sha1 ... you might think of the sha1's like URIs and the branches like DNS human readable :) 19:51:02 sometimes the pointers get out of sync ... and the error messages are not always that helpful 19:52:22 The idea is mathematically nice but it has both bugs (like that above) and the UI is not helpful. 19:53:47 Like why do do I check out the origin master but it doesn't know I am on the master branch, when for most of the time that is all many people will deal with? 19:54:20 Like you can do clone --recursive but not pull --recursive 19:56:38 true, what version of git do you have " git --version " 19:56:51 i think they are making some improvement to the recursive stuff 19:57:15 git version 1.7.6 19:57:26 oh good 20:03:15 so i just did 20:03:18 zip -r ../plugin.xpi * 20:03:31 and it installed in firefox 10 20:03:39 now to test it ... 20:03:53 perhaps i should have just done /content and /skin instead of * 20:07:38 bblfish (~bblfish@AAubervilliers-651-1-340-214.w83-114.abo.wanadoo.fr) has joined #dig 20:11:32 I just tried http://www.w3.org/ns/auth/cert.rdf 20:11:50 Content-Typeapplication/rdf+xml; qs=0.9 20:12:03 but it didnt fire up tabulator 20:13:29 Don't install a snapshot, make apointer 20:14:00 just make a one-line file in your profile's extension directory containing the full filename of the top level of the tabulator-firefox project 20:14:19 as described in the doc 20:14:56 ah 20:15:13 then you can pull new versions and just restart ff 20:15:36 I am happily categorizing tramnsactions again 20:17:03 presbrey: how did you get those error messages regarding the menu, just plain firebug? 20:17:15 oh good! 20:18:26 I got those error messages in the js console 20:18:55 I think it helps if the extension fails to work silently to run the firefox app with -jsconsole on the command line 20:19:09 got it, thanks 20:26:38 works? 20:30:06 timbl: yes thanks, js console works 20:57:44 extensions.logging.enabled;true 20:58:02 nglayout.debug.disable_xul_cache;true 20:58:15 I have a profile just for this; firefox -P tabulator 20:58:33 set my homepage to http://www.w3.org/ns/auth/acl or other non-skin'd graph 21:00:47 presbrey: thanks! 21:41:15 lkagal (~lkagal@pool-96-237-240-136.bstnma.fios.verizon.net) has joined #dig 23:06:57 presbrey has quit (Ping timeout: 260 seconds) 23:19:58 presbrey (~presbrey@2001:4830:2446:b5:aede:48ff:fe00:2001) has joined #dig 23:20:27 scor has quit (Quit: scor)