Ecoinformatics Redmine: Issueshttps://projects.ecoinformatics.org/ecoinfo/https://projects.ecoinformatics.org/ecoinfo/ecoinfo/favicon.ico?14691340362009-04-15T01:22:29ZEcoinformatics Redmine
Redmine Kepler - Bug #3986 (Resolved): RExpression confounding R working directory and .kepler folderhttps://projects.ecoinformatics.org/ecoinfo/issues/39862009-04-15T01:22:29ZOliver Soongsoong@nceas.ucsb.edu
<p>If I have two RExpression actors, I might expect them to operate in different working directories, yet still be able to pass data from one to the other. This fails, as shown in the example at the URL.</p>
<p>NOTE: The example will first fail because of another bug. To see this problem, you have to first do the workaround described in the second comment of bug 3985.</p> Kepler - Bug #3985 (Resolved): Types resolved to unacceptable typeshttps://projects.ecoinformatics.org/ecoinfo/issues/39852009-04-15T00:46:31ZOliver Soongsoong@nceas.ucsb.edu
<p>URL provides a sample workflow that has been demonstrating the error on my system. Because this problem has been confusingly inconsistent in the past, I'm not 100% convinced it will produce an error for others, but it seems to be causing problems on three builds on my system, two builds in a VM, and two systems in the vislab here, so hopefully I won't be alone this time. It's affecting both 1.0.0 and 1.x dev.</p> Kepler - Bug #3984 (Resolved): RequireVersion parameter cannot be deleted from a saved workflowhttps://projects.ecoinformatics.org/ecoinfo/issues/39842009-04-14T18:16:48ZOliver Soongsoong@nceas.ucsb.edu
<p>In an unsaved workflow (no associated file), RequireVersion can be added and deleted as expected. If I save an entirely empty workflow to a file, I can add RequireVersion but cannot delete it. The URL is an example workflow.</p> Kepler - Bug #3973 (New): Additional information for 64-bit users installing 1.0.0https://projects.ecoinformatics.org/ecoinfo/issues/39732009-04-10T23:12:04ZOliver Soongsoong@nceas.ucsb.edu
<p>Kepler 1.0.0 on 64-bit Windows systems seems to require a 32-bit java install, which is not the default provided by java.com. The error claims Java 1.5.0 could not be found, even if 1.5.0 is installed and even though it works with Java 1.6.0. At the very least, a note indicating such at kepler-project.org would help, and this seems to affect both XP and Vista.</p> Kepler - Bug #3925 (Resolved): ant run does not always work from paths with spaceshttps://projects.ecoinformatics.org/ecoinfo/issues/39252009-03-31T20:18:53ZOliver Soongsoong@nceas.ucsb.edu
<p>XP Pro x64 SP2, Java 1.6.0_11, Kepler 1.x dev build 17250</p>
<p>I checked out kepler to C:\Documents and Settings\soong\Desktop, but ant run doesn't work right. If I use the 8.3 filename, it almost works. If I create a junction at a path with no spaces, it works.</p>
<p>Path: C:\Documents and Settings\soong\Desktop\kepler\build-area<br />Fails due to whitespace quoting problem. Error says it can't find the main class "and" which it picks up from Documents and Settings.</p>
<p>Path: C:\Docume~1\soong\Desktop\kepler\build-area<br />Runs, but can't seem to access the network. I can't search KNB. There is no error, but it never returns. Existing datacans from loaded workflows are stuck at "Busy".</p>
<p>Path: C:\tmp\kepler\build-area<br />Seems to work for now.</p> Kepler - Bug #3912 (Resolved): RExpression actor has a single line text entry fieldhttps://projects.ecoinformatics.org/ecoinfo/issues/39122009-03-20T22:37:27ZOliver Soongsoong@nceas.ucsb.edu
<p>XP Pro x64 SP2, Java 1.6.0_11, Kepler 1.x dev Rev. 17126 from svn</p>
<p>The RExpression actor source entry is only a single line. It does not convert pasted newlines into in the XML.</p>
<p>Current the actor's XML has this bit:</p>
<p><property name="expression" class="ptolemy.kernel.util.StringAttribute" value="a <- c(1,2,3,5) plot(a)"><br /> <display name="R function or script"/><br /></property><br /><property name="_editorFactory" class="ptolemy.vergil.toolbox.TextEditorTableauFactory"><br /> <property name="attributeName" class="ptolemy.kernel.util.StringAttribute" value="expression"><br /> </property><br /> <property name="columnsDisplayed" class="ptolemy.data.expr.Parameter" value="80"><br /> </property><br /> <property name="rowsDisplayed" class="ptolemy.data.expr.Parameter" value="40"><br /> </property><br /></property></p>
<p>Replacing that bit of XML with the following, based on older RExpression actors, seems to work:</p>
<p><property name="expression" class="ptolemy.kernel.util.StringAttribute" value="a <- c(1,2,3,5) plot(a)"><br /> <display name="R function or script"/><br /> <property name="R Expression" class="ptolemy.actor.gui.style.TextStyle"><br /> <property name="height" class="ptolemy.data.expr.Parameter" value="20"><br /> </property><br /> <property name="width" class="ptolemy.data.expr.Parameter" value="80"><br /> </property><br /> </property><br /></property></p> Kepler - Bug #3910 (Resolved): RExpression multiport input drops record columnshttps://projects.ecoinformatics.org/ecoinfo/issues/39102009-03-19T19:27:25ZOliver Soongsoong@nceas.ucsb.edu
<p>XP Pro x64 SP2, Java 1.6.0_11, Kepler 1.0.0 from kepler-project.org</p>
<p>Sending multiple records to the RExpression actor via a multiport input does not work as expected. With a single record, it behaves as expected. With two different records, it seems to drop the last column from each. If the two records come from the same source, the actor behaves as expected. If the two records come from different sources but contain the same data, the actor behaves as expected. With three different records, it seems to drop the first and last column. If the three records come from the same source, the actor behaves as expected. If the three records come from different sources but contain the same data, the actor behaves as expected. If only two of the three records contain unique data, then the actor drops the last column from all three.</p>
<p>The URL points to a workflow demonstrating the problem.</p> Kepler - Bug #3909 (Resolved): RExpression fails with certain port nameshttps://projects.ecoinformatics.org/ecoinfo/issues/39092009-03-19T19:08:01ZOliver Soongsoong@nceas.ucsb.edu
<p>XP Pro x64 SP2, Java 1.6.0_11, Kepler 1.0.0 from kepler-project.org</p>
<p>Port names that are not simple R names causes the RExpression actor to fail with the error message that R cannot be found. The error message is misleading, and the variable names could still be used through the assign function. Escaping the name with backquotes will work in nearly all cases, but I think the assign function is slightly more robust.</p> Kepler - Bug #3901 (Resolved): unescaped & in <entity name="a & b"> on line 4https://projects.ecoinformatics.org/ecoinfo/issues/39012009-03-18T02:08:48ZOliver Soongsoong@nceas.ucsb.edu
<p>XP Pro x64 SP2, Java 1.6.0_11, Kepler 1.0.0 from kepler-project.org</p>
<p>If a workflow is saved with an & in the file name, even though the file name is valid in the file system, Kepler throws an exception because the name attribute is unescaped in the outermost TypedCompositeActor entity. If Kepler is trying to re-open the workflow similar to how a save as is handled, then it's choking while attempting to parse the file. The & should be escaped as & in the name attribute, I think.</p> Kepler - Bug #3893 (Resolved): EML 2 Dataset not working with local datahttps://projects.ecoinformatics.org/ecoinfo/issues/38932009-03-16T22:43:51ZOliver Soongsoong@nceas.ucsb.edu
<p>Kepler 1.0.0, Java 1.6.0_11, XP 64-bit</p>
<p>I downloaded an EML/XML file and some text data tables to experiment with. The EML 2 Dataset actor is having problems running with this local data.</p>
<p>This bug may be related to <a class="issue tracker-1 status-3 priority-2 priority-default closed" title="Bug: EML 2 Dataset file path confusion (Resolved)" href="https://projects.ecoinformatics.org/ecoinfo/issues/3892">#3892</a>: <br /><a class="external" href="http://bugzilla.ecoinformatics.org/show_bug.cgi?id=3892">http://bugzilla.ecoinformatics.org/show_bug.cgi?id=3892</a></p>
<p>The URL is a zip of four files: <br />EML file: judithk.609.21.xml<br />Data file: judithk.657.1-Dailyrain1980.txt<br />problem workflow: full path.xml<br />unrelated workflow: filename only.xml</p> Kepler - Bug #3892 (Resolved): EML 2 Dataset file path confusionhttps://projects.ecoinformatics.org/ecoinfo/issues/38922009-03-16T22:39:23ZOliver Soongsoong@nceas.ucsb.edu
<p>Kepler 1.0.0, Java 1.6.0_11, XP 64-bit</p>
<p>The EML 2 Dataset actor parameters EML File and Data File look for file paths. If I open a new workflow, add the EML 2 Dataset actor, and browser to two files, their full paths are recorded in the actor. I can save this workflow and open it with no problems.</p>
<p>If I place the saved workflow in the same folder as the EML and data files, open the saved workflow, edit the EML 2 Dataset actor, and browse to the same two files, then the full paths are replaced with local paths relative to the workflow's xml file (i.e., just the filename). If I save this modified workflow and try to reopen it, I get an error:</p>
<p>Data entity : Dailyrain1992.txt, failed to be downloaded and please check the data link in metadata</p>
<p>The URL is a zip of four files: <br />EML file: judithk.609.21.xml<br />Data file: judithk.657.1-Dailyrain1980.txt<br />first workflow (no error on load): full path.xml<br />second workflow (error on load): filename only.xml</p> Kepler - Bug #3890 (Resolved): EML 2 Dataset has trouble with a certain data tablehttps://projects.ecoinformatics.org/ecoinfo/issues/38902009-03-16T20:49:59ZOliver Soongsoong@nceas.ucsb.edu
<p>Kepler 1.0.0, Java 1.6.0_11, XP 64-bit</p>
<p>Kepler has difficulty with table Dailyrain2004.txt from judithk.609.21 in KNB. The workflow at the URL is simply the EML 2 Dataset actor for that table and nothing else. The error message is not particularly helpful.</p> Kepler - Bug #3807 (Resolved): reserved symbols in record names passed to the RExpression actor g...https://projects.ecoinformatics.org/ecoinfo/issues/38072009-02-04T00:08:55ZOliver Soongsoong@nceas.ucsb.edu
<p>XP Pro x64 SP2, Java 1.6.0_11-b03, Kepler 1.0.0, R 2.8.0</p>
<p>The URL contains a bugged Kepler workflow. The data from KNB contains a column called %CC that causes an error when R is executing Kepler's RExpression initialization code. This causes Kepler to think that R is not found.</p>
<p>A frustrating workaround is to disassemble the record and reassemble it, changing the name of the offending column.</p> Kepler - Bug #3803 (New): Cannot open a workflow through the command prompt, open with, or send t...https://projects.ecoinformatics.org/ecoinfo/issues/38032009-01-30T01:04:24ZOliver Soongsoong@nceas.ucsb.edu
<p>XP Pro x64 SP2, Java 1.6.0_11, Kepler 1.0.0 from kepler-project.org</p>
<p>If I right click a workflow file and select Open With->Kepler, I get an error.</p>
<p>If I add a shortcut to Kepler to my SendTo menu, then right click the workflow file and use the Send To link, I get the same error.</p>
<p>If I call Kepler from the command prompt with the workflow as the only argument, I get the same error.</p>
<p>An image of the error is at the URL.</p>
<p>I think this feature is available under linux: <br /><a class="external" href="https://dev.kepler-project.org/developers/reference/executing-kepler-from-the-command-line">https://dev.kepler-project.org/developers/reference/executing-kepler-from-the-command-line</a></p> Kepler - Bug #3801 (Resolved): open dialog, common places pane has white box instead of texthttps://projects.ecoinformatics.org/ecoinfo/issues/38012009-01-29T23:58:59ZOliver Soongsoong@nceas.ucsb.edu
<p>XP Pro x64 SP2, Java 1.6.0_11, Kepler 1.0.0 from kepler-project.org.</p>
<p>When I go to File->Open File..., the common places pane on the left has five icons and five white boxes where the icon labels are supposed to be.</p>
<p>Image at the URL.</p>