[xwiki-devs] Curriki development how-to
hello developers, I've been somewhat (but only somewhat) successfully building Curriki with mvn which, indeed, used the DB connection where I would install it. I would like to understand a bit more of the architecture... should I rather discuss it per email or open a wiki page that tries to sketch it ? Here's what I could see: - curriki is made of many modules, some of which are xwiki modules, some are own - some of the own resources are XWiki documents - the whole packaging ends up by building a web-archive and an xwiki- archive (the xwiki archive seems to contain a lot of the xwiki documents) - the build process seems to upload this web-archive to the DB (connected to a temporary curriki) it's easy then to deploy the web-archive and use. However the web-archive seems incomplete on the played. A simple one to check is Main/Board. I tried (re-)importing it but it complains that UselessPanel is not there. Three questions: - is all the above correct ? - is there another way to upload single xwiki documents ? - how are XWiki documents in source code edited ? (that's velocity in xml as far as I can see). thanks in advance paul
Hi Paul, It would clearly help if we start sketching the Curriki architecture in a Wiki page. We can of course discuss it by email also. Here are my answers/insight about what you said. Paul Libbrecht wrote:
hello developers,
I've been somewhat (but only somewhat) successfully building Curriki with mvn which, indeed, used the DB connection where I would install it.
I would like to understand a bit more of the architecture... should I rather discuss it per email or open a wiki page that tries to sketch it ?
Here's what I could see:
- curriki is made of many modules, some of which are xwiki modules, some are own
Yes. The Curriki specific modules are small plugins, a big GWT application and a skin (with macros). There are 2 plugins that will move to the core (invitationmanager and spacemanager) There is also code in the XWiki Documents to display data.
- some of the own resources are XWiki documents I would say ALL curriki resources are XWiki documents. This is if you mean "resources" being data. The XWiki documents also include a lot of code (presentation but also a but logic). - the whole packaging ends up by building a web-archive and an xwiki-archive Yes.. The xwiki archive is fully curriki specific. (the xwiki archive seems to contain a lot of the xwiki documents) - the build process seems to upload this web-archive to the DB (connected to a temporary curriki)
it's easy then to deploy the web-archive and use.
However the web-archive seems incomplete on the played. A simple one to check is Main/Board. I tried (re-)importing it but it complains that UselessPanel is not there.
What might definitively happen is that the SVN code of xwiki archive is not fully complete. The reason is that we work always from an existing Curriki install (never from scratch). So the initial wiki in the SVN might be slightly incomplete, as can be the xwiki.cfg file.
Three questions: - is all the above correct ? almost - is there another way to upload single xwiki documents ? There is an "Admin.Import" page that was made for Curriki to load the xml files from the file system. - how are XWiki documents in source code edited ? (that's velocity in xml as far as I can see).
Using the wiki or XEclipse. The Curriki team has a development server for this (which contains the production database copied over after each release). One way to solve your problem could be to give you our local development database (it's not too big). Now mine for example is not too up to date. The most up-to-date xwiki curriki database is the development server the Curriki team is working on. Basically there is work needed : - to commit all currently in development pages on the dev server in the SVN - export some key pages from the dev server in SVN that might not be up-to-date (like XWiki Prefs, Admin user, rights) - verify that the xwiki archive is a functionnal Curriki archive. Ludovic
thanks in advance
paul ------------------------------------------------------------------------
_______________________________________________ devs mailing list [email protected] http://lists.xwiki.org/mailman/listinfo/devs
-- Ludovic Dubost Blog: http://blog.ludovic.org/ XWiki: http://www.xwiki.com Skype: ldubost GTalk: ldubost
Le 11 janv. 08 à 15:25, Ludovic Dubost a écrit :
However the web-archive seems incomplete on the played. A simple one to check is Main/Board. I tried (re-)importing it but it complains that UselessPanel is not there. What might definitively happen is that the SVN code of xwiki archive is not fully complete. The reason is that we work always from an existing Curriki install (never from scratch). So the initial wiki in the SVN might be slightly incomplete, as can be the xwiki.cfg file.
But shouldn't the error be visible somewhere ? xwiki.log only says UselessPanel is missing whereas this panel is in the archive.
Three questions: - is all the above correct ? almost - is there another way to upload single xwiki documents ? There is an "Admin.Import" page that was made for Curriki to load the xml files from the file system.
URL ?
- how are XWiki documents in source code edited ? (that's velocity in xml as far as I can see). Using the wiki or XEclipse.
So nothing java-aware, right? What annoys me is that XEclipse can't open the XML files... that'd be quite nifty.
The Curriki team has a development server for this (which contains the production database copied over after each release).
One way to solve your problem could be to give you our local development database (it's not too big). Now mine for example is not too up to date. The most up-to-date xwiki curriki database is the development server the Curriki team is working on.
The work below is needed but may be big (since it is about *publication*). I wonder if at this stage I should not propose to enter an NDA to try to do this work (in a minimal form) and let the Curriki team just review the list of pages I feel are useful. paul
Basically there is work needed : - to commit all currently in development pages on the dev server in the SVN - export some key pages from the dev server in SVN that might not be up-to-date (like XWiki Prefs, Admin user, rights) - verify that the xwiki archive is a functionnal Curriki archive.
Hi Paul, Sorry for the late response.. We are in the middle of the Curriki 1.5 release. The schedule is tight.. Paul Libbrecht wrote:
Le 11 janv. 08 à 15:25, Ludovic Dubost a écrit :
However the web-archive seems incomplete on the played. A simple one to check is Main/Board. I tried (re-)importing it but it complains that UselessPanel is not there. What might definitively happen is that the SVN code of xwiki archive is not fully complete. The reason is that we work always from an existing Curriki install (never from scratch). So the initial wiki in the SVN might be slightly incomplete, as can be the xwiki.cfg file.
But shouldn't the error be visible somewhere ? xwiki.log only says UselessPanel is missing whereas this panel is in the archive.
I'm not sure what the error is here.. How do you do the import ? What exactly are you importing ?
Three questions: - is all the above correct ? almost - is there another way to upload single xwiki documents ? There is an "Admin.Import" page that was made for Curriki to load the xml files from the file system.
URL ?
Once your curriki wiki runs.. You need to get the Admin/Import page in your wiki.. Ask me on skype I can give you a text copy.. This page is in the wiki pages of the Curriki distribution.. Once it's in your wiki go to it and edit it to set the path of where you Curriki source dir of the wiki pages are.. Then you can run the import and it should import pages (it might not be all pages). The script could be changed to load all the pages in the directory.
- how are XWiki documents in source code edited ? (that's velocity in xml as far as I can see). Using the wiki or XEclipse.
So nothing java-aware, right? What annoys me is that XEclipse can't open the XML files... that'd be quite nifty.
Right.. It will in the future :) We indeed need to be able to read an xwiki xml file and load it in the wiki. Currently you need to make is a "xar" with a package.xml file to load one wiki page in the wiki.. It should not be too difficult to implement a very simple load XML script..
The Curriki team has a development server for this (which contains the production database copied over after each release).
One way to solve your problem could be to give you our local development database (it's not too big). Now mine for example is not too up to date. The most up-to-date xwiki curriki database is the development server the Curriki team is working on.
The work below is needed but may be big (since it is about *publication*). I wonder if at this stage I should not propose to enter an NDA to try to do this work (in a minimal form) and let the Curriki team just review the list of pages I feel are useful.
YES ! We can definitively help you.. Ludovic
paul
Basically there is work needed : - to commit all currently in development pages on the dev server in the SVN - export some key pages from the dev server in SVN that might not be up-to-date (like XWiki Prefs, Admin user, rights) - verify that the xwiki archive is a functionnal Curriki archive.
------------------------------------------------------------------------
_______________________________________________ devs mailing list [email protected] http://lists.xwiki.org/mailman/listinfo/devs
-- Ludovic Dubost Blog: http://blog.ludovic.org/ XWiki: http://www.xwiki.com Skype: ldubost GTalk: ldubost
participants (2)
-
Ludovic Dubost -
Paul Libbrecht