- xwiki-devs@xwiki.org - lists.xwiki.org

[xwiki-devs] about XWiki XEclipse Restification
by Jun Han 13 May '11

13 May '11

Dear all, First of all, I am very glad that I can participate in the GSOC project and thanks a lot for giving this great oppotunity. During the Community bonding period, I will look at the JIRA issues about XEclipse. However, I have several questions about XEclipse development: 1. Now the whole XWiki project is moving to github, where would XEclipse project be hosted, in SVN or git? I searched for XEclipse in github, and do not find a related entry. 2. what is the build method, using maven, Tycho or just in Eclipse IDE? This is related to XEclipse-150 issue In JIRA, which requires the whole XEclipse project to be built from maven and Tycho. I can already run XEclipse from Eclipse IDE. But if maven is to be used, I think I will learn some maven as well. Best regards Jun Han

5 12

[xwiki-devs] RPC Client Issue
by Uday Ghattamaneni 13 May '11

13 May '11

Hello, I am working on developing RPC client for Xwiki3.0 and got the below exception Failed to invoke method getObject in class com.xpn.xwiki.xmlrpc.XWikiXmlRpcApiImpl: [Page 'UserSpace.UserTestPage' cannot be accessed] I got the error when I am trying to retrieve objects as mentioned in the code below. My main problem is the code is working fine with some pages, I am getting actual values for some pages, but some pages return the above error message, I checked the privileges for those two pages, they are all same for the working one and for the page which I am getting this error. . .. ... XWikiXmlRpcClient rpc = new XWikiXmlRpcClient(url); rpc.login(username, password); XWikiObjectSummary xosum = new XWikiObjectSummary(); xosum.setClassName("XWiki.XWikiRights"); xosum.setPageId(page.getId()); XWikiObject xwo = *rpc.getObject(xosum);* ...... ... .. Also, I am having problems with the RPC client functionality I tried using Apache XML-RPC Client ,Client-side Proxy for Java and also native Swizzle API for RPC , all of them provide limited functions where all the functions listed in Remote API specifications<http://confluence.atlassian.com/display/CONFDEV/Remote+API+Specification>are not supported by any of these mainly some of such functions which I tried to implement is getPagePermissions(String token, String pageId) , I got an exception *org.codehaus.swizzle.confluence.ConfluenceException: No such handler: confluence1.getPagePermissions* while I tried to use Apache XML-RPC Client , I don't have any issue with some of the functions like getPages() etc. Is there any specific way where we could make all the functions listed in the Remote API specifications<http://confluence.atlassian.com/display/CONFDEV/Remote+API+Specification>work as specified, if any one could send me a sample working program or any document related would be helpful. Thanks in advance, Uday

2 1

[xwiki-devs] [investigation] XWiki 5 minutes Experience - Need users' help
by LD 13 May '11

13 May '11

Hi, We are launching an investigations called "XWiki 5 minutes experience" (for those who don't know, an investigation is the work before we actively work on a new development). In this case, the objective is to identify and simplify all the common things that admins and users are doing in their first five minutes of XWiki usage. The first step to this investigation is to identify all the actions they are doing and which issues they struggle with. This is where we need the help of our fellow XWiki users. Please come on this page: http://dev.xwiki.org/xwiki/bin/view/Design/XWiki5minutesExperienceMostCommo… and help us find out all the common actions (and problems with these actions) that users (actually admins and then normal users) are trying to do the first time they use XWiki. The listing of all actions and problems with be followed by studying the solutions we can implement to fix the situations reported. Some solutions might be very easy to implement and we will push then in the next release cycle, while some others will require more work and will be proposed as new work for the future release cycles. Make you voice heard as this can have a big impact on the XWiki ease of use. Ludovic

2 1

[xwiki-devs] [Performance] Livetable calls to DB
by Vincent Massol 13 May '11

13 May '11

Hi devs, Since we can now change the log levels at runtime I was trying to set "org.hibernate.SQL" to "debug". It works well ;) I reload the AllDocs page and was surprised to see that we have 5 requests to the DB: 2011-05-13 08:53:51,524 [http://localhost:8080/xwiki/bin/view/Main/AllDocs] DEBUG o.h.SQL - select top ? distinct xwikidocum0_.XWD_WEB as col_0_0_, xwikidocum0_.XWD_NAME as col_1_0_, xwikidocum0_.XWD_DATE as col_2_0_ from xwikidoc xwikidocum0_ where (xwikidocum0_.XWD_HIDDEN<>1 or xwikidocum0_.XWD_HIDDEN is null) and 1=1 and xwikidocum0_.XWD_AUTHOR='XWiki.Admin' order by xwikidocum0_.XWD_DATE desc 2011-05-13 08:53:52,134 [http://localhost:8080/xwiki/bin/get/XWiki/LiveTableResults?outputSyntax=pla…] DEBUG o.h.SQL - select list3_.XWL_VALUE as col_0_0_ from xwikiobjects baseobject0_, xwikilists dbstringli1_ inner join xwikiproperties dbstringli1_1_ on dbstringli1_.XWL_ID=dbstringli1_1_.XWP_ID and dbstringli1_.XWL_NAME=dbstringli1_1_.XWP_NAME, xwikilistitems list3_, xwikidoc xwikidocum2_ where dbstringli1_.XWL_ID=list3_.XWL_ID and dbstringli1_.XWL_NAME=list3_.XWL_NAME and baseobject0_.XWO_CLASSNAME='XWiki.TagClass' and baseobject0_.XWO_ID=dbstringli1_.XWL_ID and dbstringli1_.XWL_NAME='tags' and xwikidocum2_.XWD_FULLNAME=baseobject0_.XWO_NAME and 1=1 2011-05-13 08:53:52,136 [http://localhost:8080/xwiki/bin/get/XWiki/LiveTableResults?outputSyntax=pla…] DEBUG o.h.SQL - select list3_.XWL_VALUE as col_0_0_ from xwikiobjects baseobject0_, xwikilists dbstringli1_ inner join xwikiproperties dbstringli1_1_ on dbstringli1_.XWL_ID=dbstringli1_1_.XWP_ID and dbstringli1_.XWL_NAME=dbstringli1_1_.XWP_NAME, xwikilistitems list3_, xwikidoc xwikidocum2_ where dbstringli1_.XWL_ID=list3_.XWL_ID and dbstringli1_.XWL_NAME=list3_.XWL_NAME and baseobject0_.XWO_CLASSNAME='XWiki.TagClass' and baseobject0_.XWO_ID=dbstringli1_.XWL_ID and dbstringli1_.XWL_NAME='tags' and xwikidocum2_.XWD_FULLNAME=baseobject0_.XWO_NAME and 1=1 2011-05-13 08:53:52,137 [http://localhost:8080/xwiki/bin/get/XWiki/LiveTableResults?outputSyntax=pla…] DEBUG o.h.SQL - select top ? distinct xwikidocum0_.XWD_WEB as col_0_0_, xwikidocum0_.XWD_NAME as col_1_0_, xwikidocum0_.XWD_NAME as col_2_0_ from xwikidoc xwikidocum0_ where (xwikidocum0_.XWD_HIDDEN<>1 or xwikidocum0_.XWD_HIDDEN is null) and 1=1 order by xwikidocum0_.XWD_NAME asc 2011-05-13 08:53:52,138 [http://localhost:8080/xwiki/bin/get/XWiki/LiveTableResults?outputSyntax=pla…] DEBUG o.h.SQL - select count(distinct xwikidocum0_.XWD_FULLNAME) as col_0_0_ from xwikidoc xwikidocum0_ where (xwikidocum0_.XWD_HIDDEN<>1 or xwikidocum0_.XWD_HIDDEN is null) and 1=1 Is that the best we can do? For example there are 2 requests that come one after another that are the same (the ones starting with "select list3"). Do you know why this is so? Couldn't we do only one such request? Thanks -Vincent

1 0

[xwiki-devs] wysiwyg editor test errors
by 许凌志(Jamesxu) 13 May '11

13 May '11

Hi, I run all the test under xwiki-enterprise-test-wysiwyg again, still got 2 failures and 22 errors, I upload it to google doc: https://docs.google.com/document/pub?id=1kh5emfmiMlcEkNUKOfxyZSyj9rRV0EjR8y… There are a lot of "com.thoughtworks.selenium.Wait$WaitTimedOutException: The specified tree node, Blog, is not present!" errors, anyone can tell me why -- Best wishes, 许凌志(Jame Xu) MOE KLINNS Lab and SKLMS Lab, Xi'an Jiaotong University Department of Computer Science and Technology, Xi’an Jiaotong University

2 3

[xwiki-devs] Test for xwiki-enterprise
by 许凌志(Jamesxu) 12 May '11

12 May '11

Thanks for Marius to teach me how to run selenium test for wysiwyg editor when I fixed a bug. Today I tried to use "mvn clean install" under xwiki-enterprise-test-wysiwyg dir to fun all the tests, but results showed that there are 2 failures and 44 errors, the source codes are the newest one. Anyone can tell me why some of the test were failed? -- Best wishes, 许凌志(Jame Xu) MOE KLINNS Lab and SKLMS Lab, Xi'an Jiaotong University Department of Computer Science and Technology, Xi’an Jiaotong University

1 0

[xwiki-devs] [VOTE] Use SLF4J directly instead of inventing xwiki-commons-logging
by Vincent Massol 12 May '11

12 May '11

Hi devs, After progressing on the new logging implementation I've come to the conclusion that we don't need our own facade in xwiki-commons-logging. I started doing the facade and then hit the issue that we need a facade for the MDC feature too (Mapped Diagnostic Context) so that would be one more class to add. If you follow in the direction what will happen is that we will, in the end, completely copy the slf4j-api sources, just under a different name. So my proposal is that we standardize on the SLF4J API and use it directly without any facade. The rationale is: * The SLF4J API is in a single jar (slf4j-api.jar) which is a minimal JAR weighting 23KB (small) * The SLF4J API can be implemented over all existing logging systems. There are currently implementation for at least log4j, jdk14 logging, commons logging, logback and more. So it's ensured that if we wanted we would be able to write our own implementation of it for whatever log system that can come up. * That's one more thing that we don't have to maintain * It's very mature since the guy who wrote it (Ceki) has gone through the log4j experience Here's my +1 Thanks -Vincent

6 6

[xwiki-devs] [Proposal] Switch from Commons Logging/Log4j to SLF4J/Logback
by Vincent Massol 11 May '11

11 May '11

Hi, I'd like to propose switching from Commons Logging/Log4j to SLF4J/Logback for the following reasons: * slf4j was written to solve the issues with commons logging (memory leaks under some circumstances). http://www.google.fr/search?aq=f&sourceid=chrome&ie=UTF-8&q=slf4j+vs+common… * logback is the successor of log4j and addresses the following when compared to lgog4j: http://logback.qos.ch/reasonsToSwitch.html * New stuff happens in logback and not in log4j * I'd also like to switch to logback since it has full-fledged JMX support which allows us to easily change logging configuration at runtime (changing log levels for example): http://logback.qos.ch/manual/jmxConfig.html Note that migrating a log4j.properties file is easy: http://logback.qos.ch/translator/ In addition if a user wants to stay on log4j it would be possible and easy: simply remove the logback jar and replace it with the slf4j-log4j one. WDYT? Thanks -Vincent

6 10

[xwiki-devs] Problem with badly formed pom.xml in saaj-api-1.3.jar
by Karel Gardas 11 May '11

11 May '11

Hello, I'm trying to build xwiki enterprise from the source tree (git) I kind of succeed in having the ./xwiki-enterprise/xwiki-enterprise-web/target/xwiki-enterprise-web-3.1-SNAPSHOT.war compiled/packaged. Now I'd like to deploy it to freshly installed tomcat 7.0.12 but while tomcat starts it complains about: 2011-05-11 13:50:23,435 [Thread-2] ERROR DefaultCoreExtensionRepository - Failed to parse descriptor [jar:file:/var/tmp/xwiki/xwiki-dev-apache/apache-tomcat-7.0.12/webapps/xwiki-enterprise-web-3.1-SNAPSHOT/WEB-INF/lib/saaj-api-1.3.jar!/META-INF/maven/javax.xml.soap/saaj-api/pom.xml] org.codehaus.plexus.util.xml.pull.XmlPullParserException: expected START_TAG or END_TAG not TEXT (position: TEXT seen ...</organization>`\r\n <d... @41:5) at org.codehaus.plexus.util.xml.pull.MXParser.nextTag(MXParser.java:1095) at org.apache.maven.model.io.xpp3.MavenXpp3Reader.parseModel(MavenXpp3Reader.java:2247) at org.apache.maven.model.io.xpp3.MavenXpp3Reader.read(MavenXpp3Reader.java:4088) and the xwiki is not installed or run on it. I've unpacked the problematic saaj-api-1.3.jar file and found that the problem is probably with this line typo: 37 <organization> 38 <name>Sun Microsystems, Inc</name> 39 <url>http://www.sun.com/</url> 40 </organization>` 41 <dependencies> See "`" at the end of 40 line. Now my question is how to fix this issue properly? My guess is that this file does not come from XWiki project... Thanks! Karel

3 12

[xwiki-devs] About jira issue-5560
by 许凌志(Jamesxu) 11 May '11

11 May '11

I am trying to fix the issue-5560, thanks for the comments of Marius on the jira issue-5560 page, it made me know the reason to cause the bug. I think out a way as folloing to sovle the problem: First, I find the js code for saving wiki pages whether using shortcuts or by clicking the buttons. According to the analysis of Marius' comments, he suggest to use editor.getSourceText(onSuccess, onFailure) to get the content from the wysiwyg editor(I find when in source mode, the editor change to an plain textarea_ "<textarea class="xPlainTextEditor" style="height: 492px;"></textarea>".) I think it is a good way to get the content dirctly from wysiwyg editor, so I searched all the js file to find the editor.getSourceText method, finnally I found it in js file under xwe, it is the js code which is generated by GWT. Cause the js file under xwe aer compressed, I did not find a good way to use this method out of GWT, so anyone can give me some help to explain the usage of method editor.getSourceText(onSuccess, onFailure), and tell me how can I use it outside GWT. -- Best wishes, 许凌志(Jame Xu) MOE KLINNS Lab and SKLMS Lab, Xi'an Jiaotong University Department of Computer Science and Technology, Xi’an Jiaotong University

2 3