-<xlog server="chat.freenode.net" channel="#existdb" date="2008-03-20">

<message nick="acabre" time="02:29:42.767Z">Hey, I have a silly little question. When I use the string function twice in the sandbox, it gives me a XPST0081: No namespace defined for prefix request error. Do I need to define a namespace for string, pr prefix it?</message>

<message nick="acabre" time="02:32:42.791Z">Maybe a little context? My query is:for $object in doc('/db/sitemap/sitemap.xml')//object</message>

<message nick="Mephisto" time="02:41:53.509Z">your query didn't fully come through, try a pastebin</message>

-<message nick="acabre" time="02:43:01.031Z">

<a href="http://pastebin.com/d5bdb1b0c">http://pastebin.com/d5bdb1b0c</a>

</message>

<message nick="Mephisto" time="02:44:46.033Z">seems to run fine on the sandbox on the exist site .. doesn't return a result though</message>

<message nick="acabre" time="02:45:24.916Z">Its using a xml file that I uploaded...</message>

<message nick="acabre" time="02:46:00.585Z">Let me post the whole error message. Its kind of behind the sandbox, but I can probably copy and paste it.</message>

-<message nick="acabre" time="02:46:56.641Z">

<a href="http://pastebin.com/d1c434206">http://pastebin.com/d1c434206</a>

</message>

<message nick="Mephisto" time="02:48:58.236Z">hm, yeah .. that looks pretty bad to me</message>

<message nick="acabre" time="02:49:14.198Z">lol, I didnt think it would be useful</message>

<message nick="acabre" time="02:49:52.179Z">the same query works if I just take the raw data (eliminate one or the other calls to string() )</message>

<message nick="acabre" time="02:50:50.428Z">let me see if I can reproduce on the provided dataset</message>

<message nick="Mephisto" time="02:51:24.014Z">maybe an element you're trying to convert has a namespace prefix which is undefined in your document?</message>

<message nick="acabre" time="02:52:48.408Z">hmm, thats a good idea</message>

<message nick="acabre" time="02:55:57.515Z">doesnt look like there are any namespaces at all</message>

<message nick="acabre" time="02:56:12.876Z">... other than xmlns</message>

<message nick="Mephisto" time="02:56:56.566Z">and that one's declared?</message>

-<message nick="acabre" time="02:58:16.444Z">

<a href="http://pastebin.com/d65060763">http://pastebin.com/d65060763</a>

</message>

<message nick="acabre" time="02:58:59.903Z">but I'm generating the XML with another utility thats not eXist. If its the XML thats the problem, I dont want to bother yall with it</message>

<message nick="Mephisto" time="03:00:37.456Z">(i'm personally not a developer, but i think they're mostly asleep)</message>

<message nick="acabre" time="03:01:33.942Z">its all good. I really appreciate your help. Maybe I'll try to get the xml to validate :)</message>

<message nick="acabre" time="03:04:21.95Z">you know what... I'll just leave the values in tags, and pull them out after with perl rather than go over the generation code again... Thanks again for your help!</message>

<message nick="Mephisto" time="03:05:45.264Z">np, after a closer review, i can't point out anything obvious that may be wrong</message>

<message nick="acabre" time="03:06:47.398Z">its probably obscure. the document has to be well formed for the xpather to traverse it.</message>

<message nick="Mephisto" time="03:09:13.967Z">but you say it only happens if you use the string() function there twice?</message>

<message nick="acabre" time="03:09:56.915Z">that's how I can reproduce it, but playing around with the sandbox on exists site, its not the case.</message>

<message nick="acabre" time="03:10:36.018Z">I figured that the namespace conflict was with string(), but you are probably right that its something to do with the xml</message>

-<message nick="perig_" time="09:52:45.8Z">

wolf77 :

<a href="http://groups.google.com/group/comp.text.xml/browse_frm/thread/9ff482610fdc5fd7/d40a985ee66f9e1c?hl=fr#d40a985ee66f9e1c">http://groups.google.com/group/comp.text.xml/browse_frm/thread/9ff482610fdc5fd7/d40a985ee66f9e1c?hl=fr#d40a985ee66f9e1c</a>

</message>

<message nick="perig_" time="09:55:12.98Z">should we crosspost to other newsgroups/mailing-lists ? I did on a few french-speaking sites... but I don't like to do this : this breaks the netiquette rules ;-)</message>

<message nick="wolf77" time="09:56:43.793Z">well, yes. not sure if we should post to xmldev... it could look like product-specific marketing.</message>

<message nick="wolf77" time="09:57:16.958Z">but maybe there's some general java list we could post to as well?</message>

<message nick="wolf77" time="09:58:30.819Z">I will update the blog later today</message>

<message nick="perig_" time="09:59:24.291Z">possibly yes. Unfortunately, I have no Usenet access here. I wonder how many XML/Java newgroups we have now...</message>

-<message nick="perig_" time="10:00:36.189Z">

mybe

<a href="http://news.gmane.org/gmane.comp.java.sun.xml.general">http://news.gmane.org/gmane.comp.java.sun.xml.general</a>

</message>

<message nick="wolf77" time="10:01:48.317Z">not much traffic on there</message>

<message nick="perig_" time="10:09:56.48Z">the XQuery / XSLT MLs then ? There are regular announcements there...</message>

<message nick="wolf77" time="10:13:00.48Z">not sure as well. most people on the XQuery ML are other product developers. not our target audience really ;-)</message>

<message nick="perig_" time="10:14:42.859Z">the real problem is to know the ratio between subscribers and posters ? :-)</message>

<message nick="DiZzZz" time="10:20:54.944Z">maybe we should ask alice to apply</message>

<message nick="perig_" time="10:21:20.529Z">I'd redign then :-)</message>

<message nick="perig_" time="10:21:31.243Z">resign</message>

<message nick="DiZzZz" time="10:25:58.737Z">good morning my friends. You should be awake right now</message>

<message nick="perig_" time="10:27:14.049Z">aren't we usually ?</message>

<message nick="DiZzZz" time="10:35:27.359Z">wolf ; i did some tests on the memproc stuff ... it works as expected. just fine.....</message>

<message nick="wolf77" time="10:35:53.074Z">good to hear. if I have some time today, I will merge it into trunk.</message>

<message nick="DiZzZz" time="10:39:27.415Z">good to see the temp fragments are gone. but... for 1.2.x ...... makes it difficult to use in a production environment (i think). I mean, my simple XSD portal page..... blows the db :-(</message>

<message nick="DiZzZz" time="10:39:30.489Z">anyways</message>

<message nick="DiZzZz" time="10:39:40.523Z">wolf: yes please :-)</message>

<message nick="DiZzZz" time="10:39:48.694Z">unles I hear objecttions</message>

<message nick="DiZzZz" time="10:39:50.426Z">[object Object]</message>

<message nick="wolf77" time="10:46:09.401Z">so you say the new code should go into 1.2.1 as well?</message>

<message nick="wolf77" time="10:46:53.851Z">it probably needs more testing. I could see it in 1.2.2 though</message>

<message nick="DiZzZz" time="10:49:16.585Z">well...... hmmmmm maybe. I realize it is a significant change</message>

<message nick="DiZzZz" time="10:49:54.797Z">but when looking at a production-release .... having a simple query flooding the dbx files...... is not a good thinghy</message>

<message nick="DiZzZz" time="10:50:04.831Z">indeed not for 1.2.1</message>

<message nick="DiZzZz" time="10:50:21.834Z">which issues are open for 1.2.1 ?</message>

<message nick="DiZzZz" time="10:53:37.169Z">(while not looking at the EC stuff)</message>

<message nick="wolf77" time="10:53:38.491Z">I just need to check a few old commits and there's still a btree issue.</message>

<message nick="DiZzZz" time="10:54:19.657Z">ah ok. well it would be a good time / content for 1.2.1</message>

<message nick="DiZzZz" time="10:54:23.402Z">lunch...</message>

<message nick="DiZzZz" time="11:27:16.532Z">lunched</message>

<message nick="wolf77" time="11:59:24.04Z">has someone ever used JAI for generating thumbnails?</message>

<message nick="wolf77" time="12:06:10.031Z">sounds like a "no" ;-)</message>

<message nick="_delirium" time="12:06:19.033Z">yeah i have tried doing some stuff with JAI</message>

<message nick="_delirium" time="12:06:27.284Z">dont we have it in our images extension module?</message>

<message nick="wolf77" time="12:06:37.058Z">and was it actually faster than AWT?</message>

<message nick="_delirium" time="12:06:50.296Z">erm dunno about performance</message>

<message nick="wolf77" time="12:07:02.342Z">ah, is the extensions module code based on JAI?</message>

<message nick="wolf77" time="12:10:05.681Z">no. well, I have to try it myself then</message>

<message nick="_delirium" time="12:16:39.977Z">hmm i definetly remember doing some work with JAI</message>

<message nick="_delirium" time="12:16:56.489Z">maybe it wasnt for the extensions module - anyways its vague so I probably wouldnt be much help</message>

<message nick="_delirium" time="12:17:00.935Z">doing anything fun?</message>

<message nick="wolf77" time="12:18:24.218Z">ah, not fun really. I created a photo annotation tool for our project in Heidelberg, based on eXist. However, they now have millions of photos and the thumbnail generation code is reaching its limits ;-)</message>

<message nick="cefn" time="12:18:46.078Z">anyone know how to reset the exist database when it appears to be corrupted a) preserving data b) not preserving data</message>

<message nick="wolf77" time="12:19:20.425Z">imagemagick works quite well, but it's a pain to integrate.</message>

<message nick="wolf77" time="12:19:42.234Z">a) can you still create a backup?</message>

<message nick="wolf77" time="12:19:59.738Z">b) stop the db and delete all files from webapp/WEB-INF/data</message>

<message nick="wolf77" time="12:21:56.997Z">c) if the error is in an index, you can also stop the db, remove all .dbx and .log files except dom.dbx, collections.dbx and symbols.dbx; restart the db, open the Java client and reindex /db.</message>

<message nick="cefn" time="12:22:40.897Z">thanks wolf - will try those - also will see if I can recreate the error for a bug filing</message>

<message nick="wolf77" time="12:22:55.757Z">that would be good</message>

-<message nick="cefn" time="12:24:06.864Z">

what's the path you're using to integrate imagemagick - I use it for various fun things but mostly from the command line

<a href="http://cefn.com/blog/muybridge.html">http://cefn.com/blog/muybridge.html</a>

</message>

<message nick="wolf77" time="12:26:24.47Z">well, I tried jmagick, the Java integration. No success. I also used the ruby bindings. Worked better, but still a pain.</message>

<message nick="wolf77" time="12:27:03.644Z">Unfortunately, the server is running windows, which makes everything even more complicated. I would like to have a cross-platform solution.</message>

<message nick="_delirium" time="12:36:26.959Z">ah those brilliant windows servers!</message>

<message nick="gregory" time="14:33:42.033Z">i like the headless installation - well done!</message>

<message nick="cefn" time="14:47:19.844Z">Back now. I vaguely recollect installing command-line imagemagick on windows through cygwin one time.</message>

<message nick="_delirium" time="16:38:36.957Z">has typeswitch changed?</message>

<message nick="cefn" time="18:15:48.538Z">even after rebuild and complete wipe of data directory, still can't upload an XHTML file</message>

<message nick="cefn" time="18:16:05.981Z">'nother bug to file - 1.2-stable</message>

<message nick="perig_" time="18:18:18.031Z">tell us more before categorizing it as a bug</message>

<message nick="cefn" time="18:19:29.978Z">'k - have webdav mounted from Mac OS as admin. Drag and drop an XHTML file into the folder. Mac OS throws an 'Error Code 50' "Sorry, the operation could not be completed because an unexpected error occurred"</message>

<message nick="cefn" time="18:19:48.283Z">The server simply reports a single line "Loading mime table from file /Users/cefn/Documents/code/jamesdemo/exist/svn/stable/mime-types.xml"</message>

<message nick="cefn" time="18:20:02.983Z">No errors</message>

<message nick="perig_" time="18:20:18.654Z">even in the logs ? would be strange...</message>

<message nick="cefn" time="18:20:27.817Z">This HTML is the result of a wget, but has been put through HTML Tidy</message>

<message nick="cefn" time="18:20:35.557Z">I'll check the logs too</message>

-<message nick="cefn" time="18:25:23.889Z">

log file extract a bit big - pastebinned it

<a href="http://pastebin.com/f7d9dabdc">http://pastebin.com/f7d9dabdc</a>

</message>

<message nick="perig_" time="18:27:46.202Z">mmmh... many things warn me :-)</message>

<message nick="cefn" time="18:27:50.438Z">I wonder if it's because of Mac OS Resource Forks? </message>

<message nick="perig_" time="18:27:53.452Z">for instance : 2008-03-20 18:22:43,771 [P1-9] DEBUG (Indexer.java [fatalError]:383) - fatal error at (1,1) : Content is not allowed in prolog.</message>

-<message nick="cefn" time="18:28:05.168Z">

<a href="http://en.wikipedia.org/wiki/Resource_fork">http://en.wikipedia.org/wiki/Resource_fork</a>

</message>

<message nick="perig_" time="18:28:08.252Z">back soon, dinner</message>

<message nick="cefn" time="18:29:07.343Z">I think content not allowed in prolog I've seen before when there's a space or newline character where the parser doesn't like it</message>

<message nick="cefn" time="18:30:10.719Z">I'll file as a bug just so all the error messages are documented and all the files are there to be inspected in one Zip - nice to write off a bug with a simple explanation maybe</message>

<message nick="DiZzZz" time="18:33:11.364Z">cefn: I worked on the webdav code, and have a mac as well</message>

<message nick="DiZzZz" time="18:33:30.72Z">which release of macos?</message>

<message nick="cefn" time="18:33:33.174Z">Any thoughts from the description</message>

<message nick="cefn" time="18:33:35.106Z">Tiger</message>

<message nick="DiZzZz" time="18:33:43.808Z">pity I am on leopard</message>

<message nick="DiZzZz" time="18:34:02.774Z">well on tiger webdav worked well ; for leopard cards are different</message>

<message nick="cefn" time="18:34:15.151Z">Is there a more 'conformant' app I could try, which does WebDav</message>

<message nick="cefn" time="18:34:35.989Z">I'm assuming it's because of the content, but maybe I'm misled by my trust of Mac</message>

<message nick="DiZzZz" time="18:34:41.707Z">Finder does not access resources sequentially anymore, yielding in some locks at my place</message>

<message nick="DiZzZz" time="18:34:44.31Z">grumble</message>

<message nick="DiZzZz" time="18:35:01.323Z">a PC/webfolders works fine :-(</message>

<message nick="DiZzZz" time="18:36:00.744Z">cefn: you run trunk right?</message>

<message nick="cefn" time="18:36:10.007Z">I'm running 1.2-stable right now</message>

<message nick="cefn" time="18:36:13.171Z">could run trunk</message>

<message nick="DiZzZz" time="18:36:20.16Z">not neccesary</message>

<message nick="DiZzZz" time="18:36:25.287Z">no code changes recently</message>

<message nick="DiZzZz" time="18:37:01.897Z">are you running the default apple/sun jdk? 1.5?</message>

<message nick="cefn" time="18:37:43.834Z">yes - gonna try Goliath Webdav</message>

<message nick="DiZzZz" time="18:39:53.77Z">cefn: at least there are parse errors</message>

<message nick="cefn" time="18:39:58.697Z">Goliath can't even mount it</message>

<message nick="DiZzZz" time="18:40:08.02Z">exist won't accept these docs</message>

<message nick="cefn" time="18:40:10.183Z">I checked well-formedness before uploading</message>

<message nick="cefn" time="18:40:16.842Z">something else I need to check?</message>

<message nick="DiZzZz" time="18:40:20.797Z">in pastebin.... sory dude</message>

<message nick="DiZzZz" time="18:40:41.615Z">in pastebin ... I see content in prolog issues ; you really should check it with a binary editor</message>

<message nick="DiZzZz" time="18:40:46.672Z">write as ascii</message>

<message nick="DiZzZz" time="18:40:57.848Z">configure you editor not to write BOM header maybe</message>

<message nick="cefn" time="18:41:10.495Z">this is the output from HTML Tidy direct - maybe some options</message>

<message nick="perig_" time="18:41:12.588Z">sure, before blaming the OS, I would blame the document :-)</message>

<message nick="DiZzZz" time="18:41:26.066Z">i blame the document as well</message>

<message nick="DiZzZz" time="18:41:38.783Z">the message is clear, IMHO</message>

-<message nick="cefn" time="18:41:49.678Z">

FYI Goliath cannot connect via Webdav at all - user credentials not accepted

<a href="http://www.webdav.org/goliath/#download">http://www.webdav.org/goliath/#download</a>

</message>

<message nick="cefn" time="18:42:05.86Z">but I still have a blank admin password - straight out of the repository</message>

<message nick="DiZzZz" time="18:42:20.5Z">well if you want to try, upload document with the interactive client</message>

<message nick="DiZzZz" time="18:42:43.201Z">cefn: well I can't give you support on non-exist-apps</message>

<message nick="DiZzZz" time="18:42:50.32Z">wen there are bugs</message>

<message nick="DiZzZz" time="18:43:06.272Z">cefn: the list of 'ok' webdav apps is rather large</message>

<message nick="DiZzZz" time="18:43:14.563Z">check our documentation</message>

<message nick="cefn" time="18:43:38.145Z">not fussed about using Mac OS Native webdav especially - as long as it shows up as a filesystem I'm happy</message>

<message nick="DiZzZz" time="18:43:39.777Z">I even am able to run a 'test suite'</message>

<message nick="DiZzZz" time="18:44:05.683Z">for webdav ; so I am confident enough that exist works perry wel here</message>

<message nick="DiZzZz" time="18:44:19.371Z">bin\client.sh -l</message>

<message nick="DiZzZz" time="18:44:27.162Z">try to insert document</message>

<message nick="cefn" time="18:44:28.754Z">I get Impossible to store a resource /Users/cefn/Documents/code/jamesdemo/tradespace/Aboutag.aspx.html: org.xml.sax.SAXParseException: An element with the identifier "withinResults" must appear in the document.</message>

<message nick="cefn" time="18:44:45.937Z">Is it validating against a schema I don't know about</message>

<message nick="DiZzZz" time="18:44:46.939Z">cefn: again... it is the document</message>

<message nick="DiZzZz" time="18:46:37.879Z">probably a reference to XHTML or so</message>

<message nick="DiZzZz" time="18:47:00.39Z">maybe nekohtml plays a role here</message>

<message nick="DiZzZz" time="18:49:24.636Z">dude, goliath is pre-history software</message>

<message nick="cefn" time="18:49:46.095Z">never used it before - just the first app which came up for webdav - will check your 'compliance list'</message>

<message nick="DiZzZz" time="18:50:07.915Z">if you want to play around.... use 'cadaver' ...., that is command line.</message>

<message nick="cefn" time="18:50:49.621Z">following up on different encoding options through html tidy - ever use it? trying 'tidy -asxhtml -ascii -m'</message>

<message nick="DiZzZz" time="18:51:35.724Z">cefn: at least not myself</message>

-<message nick="cefn" time="18:52:20.144Z">

wonder if you could take a look at tidy man page - see the options I've got to work with, and maybe suggest the ones which eXist will like

</message>

<message nick="DiZzZz" time="18:52:43.195Z">well, macosx finder works well enough, the document hurts</message>

<message nick="DiZzZz" time="18:53:09.511Z">don't you have a binary editor to check your document?</message>

<message nick="DiZzZz" time="18:53:23.961Z">textmate has a binary edit mode</message>

<message nick="cefn" time="18:54:44.731Z">I don't think I understand what you mean by a binary editor, as opposed to a character editor - I'm using Oxygen to view and validate the doc - from the man page, tidy will completely rewrite the document from scratch with any character encoding you like </message>

<message nick="cefn" time="18:55:09.124Z">I can always fire up vi, but am I looking for something specific</message>

<message nick="DiZzZz" time="18:55:36.611Z">hmmmm yes and no. These BOM chars are 3 or 4 invisible chars in front of a UTF-8 document ; which give parse errors</message>

<message nick="DiZzZz" time="18:55:55.487Z">that is for the content in prolog thingy</message>

<message nick="DiZzZz" time="18:56:11.368Z">for the validation.... your actual problem.....</message>

<message nick="cefn" time="18:56:23.925Z">nothing visible in vi</message>

<message nick="DiZzZz" time="18:56:35.972Z">do you have *any* grammar references in your document?</message>

<message nick="DiZzZz" time="18:56:53.355Z">(which of the problems are we chasing now)</message>

<message nick="cefn" time="18:57:18.73Z">I was trying to see anything bad by way of characters through vi</message>

<message nick="cefn" time="18:57:37.605Z">but yes it's XHTML with full headers, DTD public etc.</message>

<message nick="DiZzZz" time="18:57:43.964Z">please send to dizzzz_at_gmail.com</message>

<message nick="DiZzZz" time="18:57:47.309Z">aaaaaaaaaaaaaaaaaah</message>

<message nick="DiZzZz" time="18:58:06.404Z">yups... it will try to resolve some of the grammar stuff</message>

<message nick="DiZzZz" time="18:58:29.296Z">org.xml.sax.SAXParseException: An element with the identifier "withinResults" must appear in the document.</message>

<message nick="DiZzZz" time="18:58:45.778Z">is a clean clear and obvious validation issue</message>

<message nick="cefn" time="18:59:14.897Z">Even with the doctype header removed the same error</message>

<message nick="DiZzZz" time="18:59:20.054Z">(although I do not understand what it means)</message>

<message nick="cefn" time