Kepler: Issueshttps://projects.ecoinformatics.org/ecoinfo/https://projects.ecoinformatics.org/ecoinfo/ecoinfo/favicon.ico?14691340362010-03-12T18:53:40ZEcoinformatics Redmine
Redmine Bug #4881 (New): RExpression outputs [false] for any matrix of booleanshttps://projects.ecoinformatics.org/ecoinfo/issues/48812010-03-12T18:53:40ZOliver Soongsoong@nceas.ucsb.edu
<p>In an RExpression, add an output port "out" and use this R script:</p>
<p>out <- matrix(c(TRUE, FALSE, TRUE, FALSE), 2)</p>
<p>I get [false] as the output token, which seems to be because RExpression.java on line 1108 says:</p>
<p>mt = new BooleanMatrixToken();</p>
<p>instead of:</p>
<p>mt = new BooleanMatrixToken(valS);</p>
<p>or something similar. I'm not actually sure such a constructor exists, but it seems like it should given the other constructors in the lines before.</p> Bug #4880 (New): RExpression cannot output matrix of stringshttps://projects.ecoinformatics.org/ecoinfo/issues/48802010-03-12T18:45:03ZOliver Soongsoong@nceas.ucsb.edu
<p>Create an RExpression with the output port "out" and for the script, use:</p>
<p>out <- matrix(letters, 2)</p>
<p>Looking at the code, it seems as if a matrix of strings was never really considered. Commas in the strings would likely cause problems, and I don't actually know if there's a StringMatrixToken.</p> Bug #4879 (New): RExpression replaces ) in string vectors with } when outputting to a porthttps://projects.ecoinformatics.org/ecoinfo/issues/48792010-03-12T18:37:49ZOliver Soongsoong@nceas.ucsb.edu
<p>In an RExpression, create the output port "out" and in the R script, use:</p>
<p>out <- c("(", ")")</p>
<p>Hooking this to a Display actor, I get {"(", "}"}. This is because RExpression.java line 1139 (I think) has:</p>
<p>temp = temp.replace(')', '}');</p>
<p>which replaces all occurrences of ), even within the string.</p> Bug #4870 (New): searching for packages with the same name from multiple sources sometimes loses ...https://projects.ecoinformatics.org/ecoinfo/issues/48702010-03-04T22:23:52ZOliver Soongsoong@nceas.ucsb.edu
<p>I have packages soong.4.14 and soong.28.5 on DEV and soong.4.15 and soong.28.9 on KNB. They all have the same title. The DEV packages require authentication to access the data but not to search. The KNB packages require authentication to search and access the data. If I search authenticated KNB or authenticated DEV individually, I see 2 entries. If I search authenticated KNB and authenticated DEV at the same time, I only see 3. I think this is caused by the way search records are "uniquely" named.</p> Bug #4869 (In Progress): changes made to workflows during dialogs before committing (Edit Paramet...https://projects.ecoinformatics.org/ecoinfo/issues/48692010-03-04T19:04:24ZOliver Soongsoong@nceas.ucsb.edu
<p>When I make changes to an actor through the dialog, it seems like the changes are made to the workflow immediately, before the "Commit" button is clicked.</p>
<p>For example, create a workflow with an EML 2 Dataset actor. Make sure Kepler has not already authenticated with KNB. Now edit the EML 2 Dataset actor to reference a data package that requires authentication. The KNB login window will immediately pop up, even though no changes have been committed. Dismiss the window in some way. Now click the Help button and close the help window. The dialog will have disappeared. Open the actor dialog again, and the change will be there.</p>
<p>I should point out that the "Cancel" button does seem to work, but I suspect (based on debugging comments and comments in code) that this acts by reverting the changes already made.</p>
<p>I've seen this for sure on EML 2 Dataset and RExpression, since changes to one trigger visual feedback and I've mucked around in the code of the other. I suspect this is a general problem, though, and it may be a result of the underlying ptolemy design.</p> Bug #4812 (New): local repository KAR [right click]->Delete fails when file does not existhttps://projects.ecoinformatics.org/ecoinfo/issues/48122010-02-19T01:18:40ZOliver Soongsoong@nceas.ucsb.edu
<p>Save a KAR to a local repository, then delete the file outside Kepler, then from the components tree right click the KAR and Delete. It will fail with an error to the console and the entry will not be removed from the tree. This could be handled better.</p> Bug #4810 (New): [right click]->Save Archive (KAR) on composite actor borks port attributeshttps://projects.ecoinformatics.org/ecoinfo/issues/48102010-02-18T21:40:38ZOliver Soongsoong@nceas.ucsb.edu
<p>Open the example MOML and save the composite actor as a KAR file. If you extract the MOML from the KAR, you'll see extra properties:</p>
<p><property name="isMultiport" value="true" class="ptolemy.kernel.util.StringAttribute"/><br /><property name="isMultiport" class="ptolemy.kernel.util.StringAttribute" value="false"><br /></property></p>
<p>and</p>
<pre><code>&lt;property name="dataType" value="unknown" class="ptolemy.kernel.util.StringAttribute"/&gt;<br /> &lt;property name="isMultiport" class="ptolemy.kernel.util.StringAttribute" value="false"&gt;<br />&lt;/property&gt;</code></pre>
<p>This causes problems.</p> Bug #4809 (New): Kepler storing authenticated search resultshttps://projects.ecoinformatics.org/ecoinfo/issues/48092010-02-18T20:18:18ZOliver Soongsoong@nceas.ucsb.edu
<p>From a clean cache, start Kepler and search using authenticated sources. Authenticate and Kepler will return search results. Restart Kepler and perform the same search again. Kepler will not ask for credentials. Try instantiating a dataset that requires authentication, and Kepler will ask for credentials. Restart Kepler and perform a search for a different term and Kepler will ask for credentials.</p>
<p>It looks like the search results are cached. Since access to new data requires authentication, new searches require authentication, and access to previously cached authenticated data does not, this may not be a problem.</p> Bug #4806 (New): add output port to Throw Exceptionhttps://projects.ecoinformatics.org/ecoinfo/issues/48062010-02-18T00:11:32ZOliver Soongsoong@nceas.ucsb.edu
<p>In some of my workflows, I have an actor that checks that certain components are installed and working before trying to run the workflow. If things are missing, it uses the Throw Exception actor to stop execution and display an error message. Right now, it's possible for the director to schedule the entire workflow before the error check, even if I use the "check" actor as a trigger. I just want some way to convince the director that needs to fire the Throw Exception actor before the rest of the workflow.</p> Bug #4801 (In Progress): out of memoryhttps://projects.ecoinformatics.org/ecoinfo/issues/48012010-02-17T22:13:10ZOliver Soongsoong@nceas.ucsb.edu
<p>ERROR: RecordingException: Unable to query data table: out of memory</p>
<p>I opened tpc01, ran it, closed it, opened tpc03-herbs, ran it, closed it, opened tpc03-large-herbivores, ran it, closed it, opened tpc03-woody, ran it, closed it, opened tpc09, ran it, and hit the above out of memory error. I have 68 runs in the wrm, but 63 of them are various iterations of the small test workflow for bug 4789. I also have those 5 Kruger KARs in a local repository and all the data for them is cached. I can start Kepler and run any of those workflows by themselves. I started Kepler with an ant run, so the jvm should have a 512MB memory max.</p>
<p>Kruger workflows: <a class="external" href="https://code.ecoinformatics.org/code/kruger/trunk/workflows">https://code.ecoinformatics.org/code/kruger/trunk/workflows</a><br />Kruger: r439<br />Kepler: wrp r23080</p>
<p>It may be worth looking at bug 4642.</p> Bug #4795 (New): RExpression & cache cleaninghttps://projects.ecoinformatics.org/ecoinfo/issues/47952010-02-13T00:15:09ZOliver Soongsoong@nceas.ucsb.edu
<p>My .kepler cache bloats pretty quickly because of RExpression's temporary files. Can we have RExpression clear it's cache folder on initialize? This way, if we need to inspect those temporary files after Kepler closes, we still can, but we'll inhibit cache bloat. I suggest doing this automatically because, while I might know what's safe to delete, I've been operating under the assumption that end users aren't expected to learn the internal structure and dependencies of .kepler.</p> Bug #4764 (New): ProvenanceRecorder.changeExecuted slow after workflow runhttps://projects.ecoinformatics.org/ecoinfo/issues/47642010-02-06T02:19:48ZOliver Soongsoong@nceas.ucsb.edu
<p>If I run any of the tpc workflows (e.g., tpc09), any subsequent changes to Kepler (say changing workflow parameters) cause Java to peg one of my CPU cores. This includes canceling changes to RExpression. I've seen this behavior on Windows XP and 7. While I haven't seen it under linux or OS X, I haven't tested those as extensively. I have tried small test workflows, and haven't seen a particularly noticeable slowdown, so it may be related to the size of the workflow run. I have to restart Kepler to get things back up to speed, and it's bad enough that I'm actually restarting Kepler after every run.</p>
<p>I'm not sure it's a memory thing. java.exe is about maxed out on memory (~0.5 GB) in the Task Manager, but the Check System Settings window says I have 46% free. I was watching jstat, and changes don't seem to trigger a flurry of garbage collection.</p> Bug #4642 (New): memory usage & slowdownshttps://projects.ecoinformatics.org/ecoinfo/issues/46422009-12-19T03:45:53ZOliver Soongsoong@nceas.ucsb.edu
<p>I just hit a big slowdown caused by OOM problems. This bug is mostly a place to put down some of the stuff I found out. I used jmap to produce histograms when Kepler was crawling and immediately after a fresh restart. When Kepler was slow, there was a single workflow open with 4 actors and the Check System Settings window. The fresh Kepler retained the wrm and cache content, but discarded the 4 actors and all the accumulated memory leaking cruft.</p>
<p>A few things jump out at me, and I'd say I'm pretty uninformed. I've formatted as Object: stale #, fresh #.</p>
<p>org.kepler.util.WorkflowRun: 39206, 29<br />javax.swing.JMenuItem: 3411, 96<br />java.util.HashMap: 689643, 22885<br />org.kepler.objectmanager.lsid.KeplerLSID: 120115, 1339<br />java.util.LinkedList: 95565, 4468<br />ptolemy.kernel.util.Location: 1837, 45</p>
<p>Interestingly enough, I have 28 wrm entries. I think something's up with the wrm, but also a lot of GUI objects seem to be hanging around as well, so there may be other things going on as well.</p>
<p>And on a side note, jps -> jmap -> jhat produces some pretty cool results.</p> Bug #4639 (New): add support for -Drev to ant change-tohttps://projects.ecoinformatics.org/ecoinfo/issues/46392009-12-18T23:18:03ZOliver Soongsoong@nceas.ucsb.edu
<p>Doesn't happen very often, but I often am not checking out the head revision, and so it would be nice if ant change-to -Dsuite=wrp -Drev=22258 worked in the logical fashion. Not at all important, but probably not very hard.</p> Bug #4613 (New): EML 2 Dataset automatically updating to latest package doesn't trigger "workflow...https://projects.ecoinformatics.org/ecoinfo/issues/46132009-12-09T01:43:12ZOliver Soongsoong@nceas.ucsb.edu
<p>I have a saved workflow that uses judithk.40.47 in an EML 2 Dataset actor. It also has "Check for latest version" checked. The LSID is urn:lsid:gamma.msi.ucsb.edu/OpenAuth/:965:34:50. judithk.40 is currently at judithk.40.48. When I open Kepler, it prompts to update the actor to use the latest version. When closing Kepler, it doesn't prompt to save changes. If I let Kepler update the workflow to use judithk.40.48 and save, the LSID is not updated.</p>