-<xlog server="chat.freenode.net" channel="#existdb" date="2008-04-28">

<message nick="wolf77" time="11:08:55.795+01:00">Hi DiZzZz</message>

<message nick="ljo" time="11:17:29.983+01:00">howdy</message>

<message nick="rebelThor" time="14:06:45.216+01:00">how does a "home collection" influences permissions? are the "owner" flags it? also, can a collection have two owners?</message>

<message nick="MoBy" time="14:11:55.487+01:00">can a collection have two owners; i guess not, that is managed through groups</message>

<message nick="MoBy" time="14:12:14.933+01:00">it is like the linux permissions</message>

<message nick="rebelThor" time="14:13:05.202+01:00">well, you can set a "home" collection for a user</message>

<message nick="rebelThor" time="14:13:26.14+01:00">what happens if you set the same collection for more than one user?</message>

<message nick="MoBy" time="14:13:41.811+01:00">i guess the last set user is the owner</message>

<message nick="rebelThor" time="14:14:53.909+01:00">i see, thanks</message>

<message nick="DiZzZz" time="17:43:07.131+01:00">hoppah!</message>

<message nick="WakkaWakka" time="19:05:51.313+01:00">hi, is there an easy way for accessing exist with as3/flex ? i know there is soap, but i am looking for somthin like xml:db</message>

<message nick="bkc" time="19:08:21.066+01:00">use the REST interface.. ?</message>

<message nick="wolf77" time="19:20:54.069+01:00">yes, I would also suggest using the REST interface. As far as I know, flex has very good support for XML-based HTTP communication.</message>

<message nick="WakkaWakka" time="19:23:51.33+01:00">i should give it a try thanks</message>

<message nick="WakkaWakka" time="19:24:08.062+01:00">is there a disadvantage over soap?</message>

<message nick="wolf77" time="19:25:03.007+01:00">in fact, REST imposes much less overhead</message>

<message nick="WakkaWakka" time="19:25:17.817+01:00">great</message>

<message nick="DiZzZz" time="19:47:14.045+01:00">Soap is is a disadvantage, if you read the literature :-)</message>

<message nick="WakkaWakka" time="19:52:40.839+01:00">okay, that's really great news. rest seams too be exactly what i want. </message>

<message nick="WakkaWakka" time="19:52:41.189+01:00">another question: i noticed that exist can also be used to store binary files. when using exist for a web application, is this much slower than leaving the files on the file system?</message>

<message nick="DiZzZz" time="19:53:44.806+01:00">Hmmmmmmm i should be about as fast... but on large documents you might expect concurrency issues</message>

<message nick="DiZzZz" time="19:54:10.611+01:00">concurrency... not that bad.... no locks or so</message>

<message nick="wolf77" time="19:54:35.204+01:00">this reminds me that Alex had a patch which effectively stores all binary resources to the filesystem as well</message>

<message nick="wolf77" time="19:54:43.205+01:00">did he merge that into trunk?</message>

<message nick="DiZzZz" time="19:55:35.736+01:00">I think he postponed until 1.3.0 / 1.2.1 is released</message>

<message nick="DiZzZz" time="19:56:06.838+01:00">I mention 1.2.1 here, because after his merge syncs from 1.3 to 1.2 would become more difficult</message>

<message nick="wolf77" time="20:21:53.759+01:00">did posting a query via the REST interface work in 1.2?</message>

<message nick="wolf77" time="20:21:59.908+01:00">it is broken in trunk</message>

<message nick="DiZzZz" time="20:22:59.599+01:00">as in..... not part _query= parameter?</message>

<message nick="DiZzZz" time="20:23:07.159+01:00">I never see it work</message>

<message nick="DiZzZz" time="20:23:16.342+01:00">I thought of adding it :-)</message>

<message nick="wolf77" time="20:23:17.043+01:00">Mmmmh</message>

<message nick="DiZzZz" time="20:23:34.727+01:00">I am wrong?</message>

<message nick="wolf77" time="20:23:47.113+01:00">as the documentation says, you can POST an XML fragment to the REST interface which describes a query to execute</message>

<message nick="DiZzZz" time="20:24:14.42+01:00">well in my XSD app / curl app I use the get construct</message>

<message nick="DiZzZz" time="20:24:28.66+01:00">can't recall the exact reason</message>

<message nick="DiZzZz" time="20:24:48.467+01:00">probably because the other did not work </message>

<message nick="DiZzZz" time="20:24:55.626+01:00">feel ashamed</message>

<message nick="DiZzZz" time="20:25:01.855+01:00">should have repaired it then</message>

<message nick="wolf77" time="20:25:41.198+01:00">well, it is Adam's code that is broken</message>

<message nick="WakkaWakka" time="20:26:33.519+01:00">what about the xml-rcp api?</message>

<message nick="DiZzZz" time="20:30:28.499+01:00">what about it?</message>

<message nick="wolf77" time="20:31:25.376+01:00">POSTing to REST is still ok in 1.2.1, but not in trunk</message>

<message nick="DiZzZz" time="20:31:37.703+01:00">surprisingly</message>

<message nick="wolf77" time="20:31:39.525+01:00">well, dinner time ...</message>

<message nick="DiZzZz" time="20:31:48.627+01:00">smakelijk!</message>

<message nick="Cassi3" time="20:54:54.74+01:00">You are developing a new version?</message>

<message nick="DiZzZz" time="21:06:11.709+01:00">1.2.1 will be the follow up of 1.2.0</message>

<message nick="DiZzZz" time="21:06:17.677+01:00">mainly bugfixes</message>

<message nick="DiZzZz" time="21:06:34.661+01:00">1.2.x is 'maintenance'</message>

<message nick="DiZzZz" time="21:08:00.988+01:00">new additions, redesign is done in trunk ; this will be called 1.3.x ; it is bleeding edge...... but less useful for production servers</message>

<message nick="DiZzZz" time="21:08:20.034+01:00">although.... I use trunk for my production servers as well :-)</message>

<message nick="Cassi3" time="21:09:27.335+01:00">illuminating</message>

<message nick="Cassi3" time="21:11:03.636+01:00">When is any of them coming out?</message>

<message nick="Cassi3" time="21:11:15.862+01:00">(an estimation)</message>

<message nick="DiZzZz" time="21:21:10.42+01:00">1.2.1 ..... maybe this week/early next week</message>

<message nick="DiZzZz" time="21:21:23.598+01:00">there is some documentation to write</message>

<message nick="DiZzZz" time="21:21:26.842+01:00">some fixes</message>

<message nick="DiZzZz" time="21:22:08.729+01:00">1.3.0 is pretty good considering "temporary fragments"</message>

<message nick="Cassi3" time="23:15:17.069+01:00">does exist favour deep nesting [11-15] or shallow [3-4]?</message>

<message nick="Cassi3" time="23:20:32.487+01:00">like in when querying what can you retrieve faster (given eXists node numbering scheme: Dynamic node numbering) the 15th nested element or the 3rd?</message>

<message nick="avanti" time="23:23:20.976+01:00">Hi. Can anyone tell me, what is the correct way in XQuery to add an element() to a "collection of elements"? Like, if I do: collection("/db/mydocs")/note and get a sequence of elements "next to each other" but want to add a custom element to the sequence.</message>

<message nick="avanti" time="23:23:39.491+01:00">I would think concat() is only for strings, no?</message>

<message nick="perig" time="23:25:41.397+01:00">retrieving a named node is very fast. Of course, decoding the id of a deeply nested node is longer than decoding an upper level one, but there shouldn't be a big difference as long as the level remains reasonnable.</message>

<message nick="perig" time="23:26:12.839+01:00">computing the relationsip should be pretty fast as well.</message>

<message nick="perig" time="23:27:16.856+01:00">avanti : there is no "collection of elements", there are collections of documents. </message>

<message nick="avanti" time="23:28:35.243+01:00">Really?? Then if I would write "let $var := <one></one><two></two><three></three>" it would throw an error, cause the elements are not nested?</message>

<message nick="perig" time="23:29:58.826+01:00">test by yourself ;-)</message>

<message nick="avanti" time="23:30:46.872+01:00">I know, sorry ;) Just thought I might ask another question really quick if the answer were yes...</message>

<message nick="perig" time="23:31:43.919+01:00">what's your problem ?</message>

<message nick="avanti" time="23:32:23.042+01:00">... but anyways, then, could I "append" an element to $docs if I had done: let $docs := collection("/db/mydocs")/note??</message>

<message nick="Cassi3" time="23:32:42.699+01:00">Thanks :)</message>

<message nick="avanti" time="23:33:44.453+01:00">I'm looping over a bunch of specific elements found within a document, but in some cases it would be incredibly practical, if I could add an extra - custom - element to the "variable" being looped.</message>

<message nick="perig" time="23:33:46.095+01:00">you wouldn't append *an* element to $docs. You would append as many elements as returned by count(collection("/db/mydocs")/note)</message>

<message nick="avanti" time="23:39:15.823+01:00">To give you the real explanation... I'm making a quite comprehensive permissions system for nested structures ("XPermissions" ;)). Every element can contain a subelement called "permissions" but when doing a check I want an ordered collection of "permissions" elements in the parent and ancestors AND in the collection (a document with root "permissions"). The thing is, in the document itself, I can just do $element/ancestor-or-self::permissions and I get a </message>

<message nick="avanti" time="23:40:22.744+01:00">... You see? I need to append this thing I've found "above" the document.</message>

<message nick="perig" time="23:41:03.159+01:00">well, it's late here :-) I'm not sure we share the same definition of collection, document, element...</message>

<message nick="avanti" time="23:42:51.065+01:00">Yeah, sorry, but to say it real short: how can I "append" something to what is returned by for instance $element/ancestor-or-self::otherelement?</message>

<message nick="avanti" time="23:43:53.88+01:00">(Actually I think we do have the same definition of collection, document and element, but it's a little too complicated to be made clear in a few lines.)</message>

<message nick="perig" time="23:45:04.056+01:00">errr... build a sequence ? i.e. ($element/ancestor-or-self::otherelement, <foo/>)</message>

<message nick="avanti" time="23:47:00.674+01:00">Yes, but then, if I loop this sequence, will it only do two iterations, cause the sequence has two instances? Or would it loop everything inside $element/ancestor-or-self::otherelement and THEN loop some more for whatever's in <foo>?</message>

<message nick="perig" time="23:48:12.502+01:00">it will loop, by definition, over the sequence. also by definition, the items in the sequence is ordered. </message>

<message nick="perig" time="23:48:28.063+01:00">are ordered</message>

<message nick="avanti" time="23:49:43.766+01:00">If you would answer a fundamental question --- would $element/ancestor-or-self::otherelement be considered a sequence or a collection??</message>

<message nick="avanti" time="23:50:04.103+01:00">... or something different?</message>

<message nick="avanti" time="23:50:31.711+01:00">Cause I do know that ((1, 2) 3) == (1, 2, 3)!</message>

<message nick="perig" time="23:50:33.303+01:00">as a result of an XPath location, it's as equence (of items)</message>

<message nick="avanti" time="23:51:00.59+01:00">Great! Then the new sequence, you told me to write will simply be one whole!</message>

<message nick="avanti" time="23:51:03.424+01:00">Thanks so much!!!</message>

<message nick="perig" time="23:51:07.7+01:00">and, indeed, it is a fundamentalanswer :-)</message>

<message nick="avanti" time="23:51:51.679+01:00">I know, I just find it really hard to find good explanations via Google besides the W3C definition... *blushing*</message>

<message nick="perig" time="23:51:57.007+01:00">to be clear : the is no idea of "sequence of sequences" in XPath 2.0</message>

<message nick="perig" time="23:53:01.775+01:00">Well, I wasn't sure because of " cause the sequence has two instances"</message>

<message nick="perig" time="23:53:46.105+01:00">let $a := (<a/>,<b/>,<c/>) return ($a, <d/>)</message>

<message nick="perig" time="23:53:52.013+01:00">4 items</message>

<message nick="avanti" time="23:53:55.247+01:00">It makes a big difference to know, that the result of an XPath is in fact a _sequence_</message>

<message nick="perig" time="23:54:27.651+01:00">yes. this is clearly the basis of XPath 2.0. </message>

<message nick="avanti" time="23:54:59.284+01:00">I'm a little too excited about my new hobby to match my present skills - but this is so powerful, every little step allows me to do so much more.</message>

<message nick="avanti" time="23:55:14.224+01:00">Please, enough... ;) ;) ;)</message>

<message nick="perig" time="23:55:21.154+01:00">I still haven't found the limits of XQuery :-)</message>

<message nick="avanti" time="23:55:41.922+01:00">What projects have you done?</message>

<message nick="perig" time="23:55:47.91+01:00">Can't workwith XSLT any more :-)</message>

<message nick="perig" time="23:56:11.052+01:00">My main project is... eXist :-)</message>

<message nick="avanti" time="23:56:17.5+01:00">Of course...</message>

<message nick="Cassi3" time="23:56:50.085+01:00">I find writing HTML with Xquery easier than with XSLT too.</message>

<message nick="avanti" time="23:56:59.688+01:00">XSLT comes in handy when wanting to convert stuff to XHTML, I would guess...</message>

<message nick="perig" time="23:59:13.419+01:00">good night</message>

<message nick="avanti" time="23:59:23.453+01:00">In reality I should think XQuery is limitless - but only due to the extensibility with Java or other.</message>

<message nick="avanti" time="23:59:33.076+01:00">Night!</message>

</xlog>