Joern Turner | 24 Jul 16:12 2014
Picon

help us spreading the word

Hi,

ever wondered how you can give back to the project?

Here's your chance ;)

I've attached a first version of an exstdb button you can put on your homepage and link to existdb.org. It will work best on light backgrounds - if you have a rather darkish hompepage i'm happy to provide further versions.

Also of course any ideas to spread the word for exist are highly appreciated.

Ah, not to forget - the image is an svg so scale it as you like to fit your page.

Thanks,

Joern
------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open
filtered | 24 Jul 12:10 2014
Picon

UI stale after triggering full backup (2.2RC1)

Hi there,

as part of the RC1->RC2 migration I wanted to start with a full backup.
I triggered the backup through the dashboard but the UI became unresponsive and showed a spinning pointer for minutes. I could not find anything related in the log and no ZIP files
on the filesystem!?

Andreas
------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open
Hendricks Trevor | 23 Jul 19:14 2014

Importing files from local file system

I can't recall how to do this and am looking for help. I need to import a file from the local system to eXist. The
server is a Windows box and the following are the basics.

I'm trying to import a csv (say import.csv) from D:\temp directory to /db/temp collection.

Xmldb:store('/db/temp', 'import.csv', 'D:\temp\import.csv') isn't working. I know my issue is in the
third parameter but I don't recall the proper format. Any help is appreciated.

Trevor Hendricks 
 (A)bort (R)etry (I)nfluence with large hammer.

------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
wolfgang | 23 Jul 17:03 2014

eXist-db 2.2RC2

A second release candidate for 2.2 is now available for download. Details can be found in the announcement:


This is a last test before 2.2 final is released. This version has huge improvements over RC1 and has proven to be quite mature and robust. We should thus not wait long to make it final. The goal is to release 2.2 final in the first half of August. 

Wolfgang


------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open
wolfgang | 23 Jul 16:26 2014

Extending documentation in the function docs

Because documentation was just mentioned in another thread, I’d like to point out that the new version of
the XQuery function documentation app provides a facility to attach additional, free-text
documentation to a function. The problem: the function signatures and documentation of the XQuery
function libraries implemented in Java is inside the Java class, which makes it hard to write a longer
description or provide examples. I have extended the function documentation app to search for
additional documentation files and display them below the signature and description automatically
extracted from the Java class. I chose markdown as the format because it is directly supported by github
and should make it easy for other people to contribute.

An example for an extended documentation page can be found here: 

http://exist-db.org/exist/apps/fundocs/view.html?uri=http://exist-db.org/xquery/util&function=util:index-keys&arity=5&location=java:org.exist.xquery.functions.util.UtilModule

The convention for additional documentation files is as follows:

* they should be stored in a collection below /db/apps/fundocs/data/docs/. The name of the collection
should correspond to the default namespace prefix of the XQuery module, e.g. "util".
* the name of the documentation file for each function follows the pattern: FunctionName_Arity.md, so if
you would like to document the function util:index-keys#5, it should go into "util/index-keys_5.md"

The markdown parser is largely compatible with the github variant of markdown. See the following page in
the wiki: 

http://exist-db.org/exist/apps/wiki//Markdown

I’ll be happy to hear your comments, suggestions. I also plan to extend the app to allow additional
documentation on the module level.

Wolfgang 
------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
Jens Østergaard Petersen | 23 Jul 10:19 2014
Picon

Re: suggestion to move to stackoverflow

To keep the same questions from popping up again and again is indeed the strong point of SO, but the Documentation App should take care of most of these.

More generally, we have several sources of eXist-db knowledge,

The Documentation App
The eXist Wiki
The Java Docs
Adam & Erik's Book
XML Prague Presentations
YouTube videos
The XQuery Wikibook
StackOverflow Q &A
Various blogs
GitHub Issues

In my opinion, there should be a concentrated effort to improve the Documentation App, to bring it up to date by reference to the other sources (I am not saying that the Book should be pillaged for this - its way of presenting the material is anyway quite different, but to neglect it would be odd as well).

To take a few examples: there are a lot of questions involving how to monitor the database, and there is a splendid new app for this, but the Documentation App is not up to date on this (nor is there any help inside the monex itself). There have been major changes in Lucene, but this is not reflected anywhere. Also, apps are introduced in the Package Manager without explanations about how they are to be installed or used.

I know there have been calls for the improvement of the Documentation App in the past, and that nothing much have come of them, but I still think this should be a priority.

Jens

On 22 Jul 2014 at 18:00:42, Joern Turner (joern.turner <at> gmail.com) wrote:

Ok, probably i've not formulated that precisely enought in the beginning. I didn't wanted to suggest to close down the list completely. Of course there needs to be a place for discussions. But i still insist on the idea that it would be better to more clearly separate questions from discussions. A mailinglist is not an effective medium if looking for answers and thus the same questions will pop up again and again.

There's nothing preventing that certain questions will evolve into discussions and a strict separation won't be enforceable anyway but IMO it would be a good thing to say - if you got a question go to stackoverflow first.


On Tue, Jul 22, 2014 at 5:53 PM, <wolfgang <at> exist-db.org> wrote:
I agree with Chris: the mailing list is not just a forum for posting questions and getting answers. As developer, the list gives me an idea of the things people are trying to achieve with eXist, the problems they face, the interests they have. Nearly all feature developments and bug fixes start with a discussion on this list. Users may not be aware of it, but we frequently change development priorities based on what people write. Seriously, even core features like the support for XQuery started with somebody telling us: hey, this is cool, we need it.

For example, take the discussion about HTML5: until recently, I personally did not see why eXist should deal with non-well-formed HTML5 at all? It was only because people insisted that I recognised it as something which cannot be ignored.

It is this social aspect which is missing on stack overflow. On the mailing list, important discussions often start with a simple question, then evolve, move on to other topics, just to come back later. We need this free flow of communication.

I don’t mind if people ask questions on stack overflow. They are taken care of as far as I can see. In fact, whenever I get a notification from stack overflow and go there, the question has already been answered and there’s no need for me to step in ;-) That’s great.

Wolfgang

------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open
------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open
Ihe Onwuka | 22 Jul 23:01 2014
Picon

Re: suggestion to move to stackoverflow

Of course there are other objectives militating in favour of an SO migration. To parallel most of my SO interaction has been in relation to freebase and insofar as getting questions answered is concerned their mailing list will get you more and quicker answers.


On Tue, Jul 22, 2014 at 4:59 PM, Joern Turner <joern.turner <at> gmail.com> wrote:
Ok, probably i've not formulated that precisely enought in the beginning. I didn't wanted to suggest to close down the list completely. Of course there needs to be a place for discussions. But i still insist on the idea that it would be better to more clearly separate questions from discussions. A mailinglist is not an effective medium if looking for answers and thus the same questions will pop up again and again.

There's nothing preventing that certain questions will evolve into discussions and a strict separation won't be enforceable anyway but IMO it would be a good thing to say - if you got a question go to stackoverflow first.


On Tue, Jul 22, 2014 at 5:53 PM, <wolfgang <at> exist-db.org> wrote:
I agree with Chris: the mailing list is not just a forum for posting questions and getting answers. As developer, the list gives me an idea of the things people are trying to achieve with eXist, the problems they face, the interests they have. Nearly all feature developments and bug fixes start with a discussion on this list. Users may not be aware of it, but we frequently change development priorities based on what people write. Seriously, even core features like the support for XQuery started with somebody telling us: hey, this is cool, we need it.

For example, take the discussion about HTML5: until recently, I personally did not see why eXist should deal with non-well-formed HTML5 at all? It was only because people insisted that I recognised it as something which cannot be ignored.

It is this social aspect which is missing on stack overflow. On the mailing list, important discussions often start with a simple question, then evolve, move on to other topics, just to come back later. We need this free flow of communication.

I don’t mind if people ask questions on stack overflow. They are taken care of as far as I can see. In fact, whenever I get a notification from stack overflow and go there, the question has already been answered and there’s no need for me to step in ;-) That’s great.

Wolfgang


------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open


------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open
Joern Turner | 22 Jul 15:03 2014
Picon

suggestion to move to stackoverflow

Hi *,

this might have been suggested earlier by others but i'd like to revive this debate:

wouldn't it be better to make stackoverflow our main forum and replace exist-open

stackoverflow is the de-facto google for developers. Everybody looking for an answer in technical questions will visit this site or quickly ending up there (coming via google).

The voting feature gives you some indication about the quality of answers (and questions) which a maling cannot offer.

Furthermore you always have to register to a mailinglist - a hurdle that is a bit counterproductive  and esp. if you are new to a product you'll hesitate to register for another inbox stuffing list that you probably don't want. IMO the hurdles to get an answer should be as low as possible.

Opinions?

Joern


------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open
Christian Schwaderer | 22 Jul 13:52 2014
Picon

XQuery performance via PHP: eXist vs Saxon

Hi all,

I have a web application based on XML data. In the current design, all user interaction is transformed via PHP to an XPath expression which becomes part of a XSLT stylesheet. The XSLT stylesheet produces an XML result document which is thereafter transformed into HTML. All the XSLT processing is done with Saxon/C 0.31 (based on Saxon 9.5.1.3 HE).

I've heard from several people that eXist might be a better solution for me. So, I've installed eXist (2.2RC1 Build:20140218) and did some tests. I was especially interested in whether I would gain faster performance by replacing my Saxon-XSLT-solution with eXist-XQuery.

But that doesn't seem to be the case. In all me tests eXist was many times slower than eXist.

One example is this XQuery:

xquery version "1.0";
for $text_of_speaker in doc("/db/apps/demo/data/hamlet.xml")//SPEAKER/text()
return concat($text_of_speaker,':',count(doc("/db/apps/demo/data/hamlet.xml")//SPEAKER[text() = $text_of_speaker]))

eXist takes about 4.8 seconds to do that query with this code:

<?php
       $time_start = microtime(true);

    include ('eXist.php');

        $db = new eXist();

        $db->connect();

        $query = 'xquery version "1.0"; for $text_of_speaker in doc("/db/apps/demo/data/hamlet.xml")//SPEAKER/text() return concat($text_of_speaker,\':\',count(doc("/db/apps/demo/data/hamlet.xml")//SPEAKER[text() = $text_of_speaker]))';

        $result = $db->xquery($query);
       
        foreach ( $result["XML"] as $xml)
         { echo $xml . ' '; }
       
        $db->disconnect();

        $time_end = microtime(true);
        $time_needed = $time_end - $time_start;

        echo "<br />Time elapsed: " . $time_needed;
?>

Saxon is able to do it in about 0.9 seconds, which is less than one fifth. Here is the code I'm using:

<?php
        $time_start = microtime(true);

        $xslt_processor_instance = new SaxonProcessor();
    
        $xslt_processor_instance->setSourceFile('/home/cis/eXist-db/webapp/WEB-INF/data/expathrepo/demo-0.2.9/data/hamlet.xml');
       
        $xslt_processor_instance->setQueryContent('xquery version "1.0"; for $text_of_speaker in //SPEAKER/text() return concat($text_of_speaker,\':\',count(//SPEAKER[text() = $text_of_speaker]))');
        $result = $xslt_processor_instance->queryToString();
        echo $result;
       
        $time_end = microtime(true);
        $time_needed = $time_end - $time_start;

        echo "<br />Time elapsed: " . $time_needed;
?>

BTW: The results are the same of course.


Is there anything wrong in my test design? Can I do anything to make eXist faster?


Otherwise, with this performance difference being the case I see no reason at all to use eXist.

Best,
Christian
------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open
Loren Cahlander | 22 Jul 05:41 2014
Picon

Trying to set CDATA for some output


I am trying to set CDATA for the script and style tags in an SVG outout.  I have put the following in the script:

declare default element namespace "http://www.w3.org/2000/svg";

declare namespace output="http://www.w3.org/2010/xslt-xquery-serialization";
declare option output:cdata-section-elements "*:script *:style";
declare option output:method "xml";
declare option output:media-type "image/svg+xml";

It does not output the CDATA in the selected nodes.  What am I missing?

------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open
Loren Cahlander | 22 Jul 05:40 2014
Picon

Trying to set CDATA for some output


I am trying to set CDATA for the script and style tags in an SVG outout.  I have put the following in the script:

declare default element namespace "http://www.w3.org/2000/svg";

declare namespace output="http://www.w3.org/2010/xslt-xquery-serialization";
declare option output:cdata-section-elements "*:script *:style";
declare option output:method "xml";
declare option output:media-type "image/svg+xml";

It does not output the CDATA in the selected nodes.  What am I missing?

------------------------------------------------------------------------------
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code with a free copy of Black Duck
Code Sight - the same software that powers the world's largest code
search on Ohloh, the Black Duck Open Hub! Try it now.
http://p.sf.net/sfu/bds
_______________________________________________
Exist-open mailing list
Exist-open <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/exist-open

Gmane