[rkward-cvs] SF.net SVN: rkward: [1491] trunk/rkward/rkward/pages/ rkward_trouble_shooting.rkh

[sjar at users.sourceforge.net]

Revision: 1491
          http://svn.sourceforge.net/rkward/?rev=1491&view=rev
Author:   sjar
Date:     2007-02-26 14:04:05 -0800 (Mon, 26 Feb 2007)

Log Message:
-----------
New section "I think I did something wrong? Where is 'Undo' and 'Redo'?".
Maybe better in RKWard for (new) Users?

Modified Paths:
--------------
    trunk/rkward/rkward/pages/rkward_trouble_shooting.rkh

Modified: trunk/rkward/rkward/pages/rkward_trouble_shooting.rkh
===================================================================
--- trunk/rkward/rkward/pages/rkward_trouble_shooting.rkh	2007-02-26 21:56:46 UTC (rev 1490)
+++ trunk/rkward/rkward/pages/rkward_trouble_shooting.rkh	2007-02-26 22:04:05 UTC (rev 1491)
@@ -22,7 +22,7 @@
 	<section title="RKWard is running since ages. Has it hung up?" id="hung_up_problems">
 		This depends. Some test and plots just need some time if you have huge amounts of data or sophisticated calculations. Therefore patience is sometimes needed.
 
-		But some statistical test are rather memory and stack hungry and can make the entire system unresponsive, if there is a lot of swapping. This is neither a bug in RKWard (most likely at least) nor in R but due to the statistical test itself.
+		But some statistical test are rather memory and stack hungry and can make the entire system unresponsive, if there is a lot of swapping. Please refere to <link href="http://cran.r-project.org/doc/Rnews/Rnews_2001-1.pdf">RNews 2001/1, pg. 10</link> for more information. Therefore, this is neither a bug in RKWard (most likely at least) nor in R but due to the statistical test itself.
 
 		An example of very memory hungry tests are the <link href="rkward://rhelp/wilcox.test">Wilcoxon Rank Sum and Signed Rank Tests</link>.
 
@@ -30,4 +30,11 @@
 
 		Therefore it's important you know the test. Moreover it's very important to save the workspace and other sections you worked on before you do such operations. There is currently no recovery functionality within RKWard in case of a crash! You can find build-in tests from R to address this memory issue but these work on Windows(TM) platforms only (see: <link href="rkward://rhelp/memory.size">Memory Size</link> and <link href="rkward://rhelp/limit">Memory Limit</link>). Sorry to say, this means there is no way for the RKWard developers to build a routine which does a "memory check" before a test is run.
 	</section>
+	<section title="I think I did something wrong? Where is 'Undo' and 'Redo'?" id="undo_redo_problems">
+		Right now RKWard doesn't have any "undo" or "redo" functionality. Therefore it's important that you know what you do before you start. It's recommended to use copies of original data and to give everything a meaningful name.
+
+		However, RKWard tries to give you as much information as possible about things you have done. For example all test and plots show the specific settings in combination with the creation date. Moreover you can check and copy any code generated for your documentation purposes
+
+		Depending on available programmer resources, the release of KDE4 and other factors it's likely to be seen in future releases.
+	</section>
 </document>


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.