Kepler: Issueshttps://projects.ecoinformatics.org/ecoinfo/https://projects.ecoinformatics.org/ecoinfo/ecoinfo/favicon.ico?14691340362012-08-31T20:49:13ZEcoinformatics Redmine
Redmine Bug #5703 (Resolved): GUI bug - Deleting an Actor with connected port causes a diamond relation t...https://projects.ecoinformatics.org/ecoinfo/issues/57032012-08-31T20:49:13ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>In the last week or so I've begun seeing a few GUI anomalies on trunk. One is:</p>
<p>Deleting an actor with a connected port causes a diamond relation to be created. If you undo the delete, the deleted actor returns, but the relation from actor1 to actor2 now connects via the diamond relation.</p>
<p>Replicate:<br />Drag out a Constant and a Display.<br />Connect output of Constant to input of Display.<br />Delete Display<br />Bug: diamond relation appears on canvas</p>
<p>This bug occurs in both kepler and vergil.</p> Bug #5663 (Resolved): KeplerGraphFrame has much duplicated codehttps://projects.ecoinformatics.org/ecoinfo/issues/56632012-08-06T23:55:37ZChristopher Brookscxh@eecs.berkeley.edu
<p>I looked in to why KeplerGraphFrame has so much duplicated code<br />and it looks like Comad uses KeplerGraphFrameExtension</p>
<p>From the 2010-11-04 comment that added this class: </p>
<pre><code>// "Added an extension for a different Drag and Drop behavior. <br /> // Currently this is only activated in Comad. Once an Actor is <br /> // dragged from the Library over another actor on the canvas it <br /> // will replace this actor and Ports are reconeected. Also <br /> // Parameter values are used for parameters of the new actor if <br /> // they have the same name. <br /> // <br /> // If an actor is dragged over a link this actor will be inserted <br /> // there. Currently this only works in Comad, where all actors <br /> // have an "input" and one "output" port."</code></pre>
<p>The right thing is to add these extension points to the ptII repository<br />and remove the duplicated code.</p> Bug #5607 (Resolved): constant actor icon won't automatically re-size since 2.3.https://projects.ecoinformatics.org/ecoinfo/issues/56072012-05-14T18:44:15Zjianwu jianwujianwu@sdsc.edu
<p>For actor 'Constant' and 'String Constant', Kepler GUI used to automatically replace long text with its first a few characters followed by '...'. I validated it in Kepler 2.2. But since Kepler 2.3 (also in Kepler trunk), these two actors won't re-size anymore. All the text in their value parameter is displayed in GUI. It is very unfriendly when the text is very long.</p>
<p>I attached a demo workflow and its displays in Kepler 2.2 and 2.3.</p> Bug #5587 (Resolved): long parameter values get replaced with scrollbarhttps://projects.ecoinformatics.org/ecoinfo/issues/55872012-04-16T22:45:46ZDaniel Crawldanielcrawl@gmail.com
<p>In the actor parameters dialog, if a parameter value has more characters than can fit in the text box, the text box is replaced with a scrollbar.</p> Bug #5573 (Resolved): remove ptolemy vendors/https://projects.ecoinformatics.org/ecoinfo/issues/55732012-02-03T20:44:53ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Remove ptolemy/src/vendors directory.</p> Bug #5571 (Resolved): user request: show execution time in runtime windowhttps://projects.ecoinformatics.org/ecoinfo/issues/55712012-02-02T20:19:00ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>User requests that workflow execution time be shown in Runtime Window. I suggest it also be shown at the bottom of each workflow window.<br /><a class="external" href="http://lists.nceas.ucsb.edu/kepler/pipermail/kepler-users/2012-February/002131.html">http://lists.nceas.ucsb.edu/kepler/pipermail/kepler-users/2012-February/002131.html</a></p> Bug #5550 (Resolved): Windows installer should be a .exe, not a .jarhttps://projects.ecoinformatics.org/ecoinfo/issues/55502011-11-22T19:41:16ZChristopher Brookscxh@eecs.berkeley.edu
<p>Downloading and running a .jar file under Windows Server 2008 is tricky.<br />Basically, the user needs to run a command from a command line, which can be<br />tricky if there are spaces or parentheses in the file name.</p>
<p>There have been questions about starting the jar file.</p>
<p><a class="external" href="https://kepler-project.org/users/faq#under-windows-when-i">https://kepler-project.org/users/faq#under-windows-when-i</a> says:</p>
<p>--start--<br />Under Windows, when I double click on the installer jar file, the installer does not start. What do I do?</p>
<p>Under Windows, the Kepler installer jar file should start the Kepler installer using Java, it should not unzip.<br />If it unzips, then see <a class="external" href="http://www.wikihow.com/Run-a-.Jar-Java-File">http://www.wikihow.com/Run-a-.Jar-Java-File</a></p>
<p>or</p>
<p><a class="external" href="http://stackoverflow.com/questions/394616/running-jar-file-in-windows">http://stackoverflow.com/questions/394616/running-jar-file-in-windows</a><br />--end--</p>
<p>Rather than shipping a .jar file, we should ship a .exe for the installer.<br />One way would be to use Launch4J, see <a class="external" href="http://launch4j.sourceforge.net/">http://launch4j.sourceforge.net/</a></p> Bug #5546 (Resolved): SDF director iterations parameter default of 0 is unfriendlyhttps://projects.ecoinformatics.org/ecoinfo/issues/55462011-11-18T00:50:15ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>I know this was debated/changed at least once in the past, but I can't remember many of the details...at one point Aaron change the default to 1, but I think did so in a way that affected existing models. <br />If we could change SDF's iterations default to 1 without affecting existing models, we'd simply the common use case and remove an extremely common stumbling block for new users.</p> Bug #5457 (Resolved): Mac OS X: Kepler.app should display module error messages in a windowhttps://projects.ecoinformatics.org/ecoinfo/issues/54572011-08-14T16:24:41ZChristopher Brookscxh@eecs.berkeley.edu
<p>When I double click on the Kepler.app icon, Kepler appears<br />to quickly start up and then exit.</p>
<p>Looking at the Console app, I see:</p>
<p>8/14/11 9:19:08.748 AM [0x0-0xe60e6].org.kepler.build.runner.Kepler: JVM Memory = 5m 512m<br />8/14/11 9:19:08.749 AM [0x0-0xe60e6].org.kepler.build.runner.Kepler: Error: The following modules are missing: <br />8/14/11 9:19:08.749 AM [0x0-0xe60e6].org.kepler.build.runner.Kepler: kepler-2.2.0<br />8/14/11 9:19:23.597 AM [0x0-0xe90e9].org.kepler.build.runner.Kepler: JVM Memory = 5m 512m<br />8/14/11 9:19:23.598 AM [0x0-0xe90e9].org.kepler.build.runner.Kepler: Error: The following modules are missing: <br />8/14/11 9:19:23.598 AM [0x0-0xe90e9].org.kepler.build.runner.Kepler: kepler-2.2.0</p>
<p>I renamed ~/.kepler to ~/.kepler.bak and the same thing occurred.</p>
<p>Renaming ~/KeplerData to ~/KeplerData.bak worked around the problem.</p>
<p>There are two bugs here:<br />1) When I click on Kepler.app and get an error, I should see a console<br />window.<br />2) Why does the contents of my KeplerData directory matter?</p> Bug #5448 (Resolved): cannot start Kepler using non-IcedTea OpenJDKhttps://projects.ecoinformatics.org/ecoinfo/issues/54482011-08-04T16:59:48ZDaniel Crawldanielcrawl@gmail.com
<p>Christopher Brooks reports:</p>
<p>Under RedHat, I was able to run Kepler-2.2.0 under Sun JDK 1.6.0_21</p>
<p>However, as expected, OpenJDK 1.6.0 fails</p>
<p>bash-3.2$ java <del>version<br />java version "1.6.0" <br />OpenJDK Runtime Environment (build 1.6.0-b09)<br />OpenJDK 64-Bit Server VM (build 1.6.0-b09, mixed mode)<br />bash-3.2$ ./kepler.sh<br />JVM Memory = 5m 512m<br /> [null] Checking for patches...<br /> [null] Building Kars...<br /> [null] Ran additional initialization for module core from class<br />org.kepler.module.core.Initialize<br /> [null] Ran additional initialization for module gui from class<br />org.kepler.module.gui.Initialize<br /> [null] X connection to localhost:10.0 broken (explicit kill or server<br />shutdown).<br />Exception in thread "main" Java returned: 1<br /> at org.kepler.build.modules.ModulesTask.execute(ModulesTask.java:106)<br /> at org.kepler.build.runner.Kepler.main(Kepler.java:94)<br />Caused by: Java returned: 1<br /> at org.kepler.build.modules.ModulesTask.execute(ModulesTask.java:106)<br /> at org.kepler.build.runner.Kepler.run(Kepler.java:218)<br /> at org.kepler.build.modules.ModulesTask.execute(ModulesTask.java:102)<br /> ... 1 more<br />Caused by: Java returned: 1<br /> at org.apache.tools.ant.taskdefs.Java.execute(Java.java:108)<br /> at org.kepler.build.Run.runSuite(Run.java:309)<br /> at org.kepler.build.Run.run(Run.java:213)<br /> at org.kepler.build.modules.ModulesTask.execute(ModulesTask.java:102)<br /> ... 3 more<br />--</del> Nested Exception ---<br />Java returned: 1<br /> at org.kepler.build.modules.ModulesTask.execute(ModulesTask.java:106)<br /> at org.kepler.build.runner.Kepler.run(Kepler.java:218)<br /> at org.kepler.build.modules.ModulesTask.execute(ModulesTask.java:102)<br /> at org.kepler.build.runner.Kepler.main(Kepler.java:94)<br />Caused by: Java returned: 1<br /> at org.apache.tools.ant.taskdefs.Java.execute(Java.java:108)<br /> at org.kepler.build.Run.runSuite(Run.java:309)<br /> at org.kepler.build.Run.run(Run.java:213)<br /> at org.kepler.build.modules.ModulesTask.execute(ModulesTask.java:102)<br /> ... 3 more<br />--- Nested Exception ---<br />Java returned: 1<br /> at org.apache.tools.ant.taskdefs.Java.execute(Java.java:108)<br /> at org.kepler.build.Run.runSuite(Run.java:309)<br /> at org.kepler.build.Run.run(Run.java:213)<br /> at org.kepler.build.modules.ModulesTask.execute(ModulesTask.java:102)<br /> at org.kepler.build.runner.Kepler.run(Kepler.java:218)<br /> at org.kepler.build.modules.ModulesTask.execute(ModulesTask.java:102)<br /> at org.kepler.build.runner.Kepler.main(Kepler.java:94)<br />bash-3.2$</p>
<p>It is unfortunate that command that was executed is not returned, nor is the<br />error message.</p>
<p>Ptolemy II, when run outside of Kepler, starts up fine with OpenJDK . . .</p>
<p>On sisyphus.eecs.berkeley.edu, we are running:</p>
<p>[bldmastr@sisyphus test]$ rpm -qa redhat-release<br />redhat-release-5Client-5.4.0.3<br />[bldmastr@sisyphus test]$ cat /etc/redhat-release <br />Red Hat Enterprise Linux Client release 5.4 (Tikanga)</p> Bug #5266 (Resolved): Two empty boxes inside Sinewave actor and other icon problems for attributeshttps://projects.ecoinformatics.org/ecoinfo/issues/52662011-01-06T00:55:56ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>While testing <a class="external" href="http://bugzilla.ecoinformatics.org/show_bug.cgi?id=5146">http://bugzilla.ecoinformatics.org/show_bug.cgi?id=5146</a> , this bug was found.<br />From the Reporting suite, if you drag a Sinewave actor to canvas, and then right-click Open Actor to see the insides of the composite, you'll see two blank boxes in this workflow. One is a ptolemy.kernel.util.Attribute (for the text that should be appearing - "Generate a sine wave."), and another is a ptolemy.kernel.util.SingletonAttribute. This problem starts happening when you're using the Reporting module.</p> Bug #5240 (Resolved): NameDuplicationException from SequencePlotterhttps://projects.ecoinformatics.org/ecoinfo/issues/52402010-11-16T19:08:16ZDaniel Crawldanielcrawl@gmail.com
<p>Reproduce with attached workflow.</p>
<p>ptolemy.kernel.util.IllegalActionException: Error creating effigy and tableau<br /> in .plot-error.a plot<br />Because:<br />Attempt to insert object named "plot" into container named ".configuration.directory.effigy2", which already contains an object with that name.<br /> at ptolemy.actor.lib.gui.Plotter.initialize(Plotter.java:167)<br /> at ptolemy.actor.lib.gui.SequencePlotter.initialize(SequencePlotter.java:130)<br /> at ptolemy.actor.process.ProcessThread.run(ProcessThread.java:135)<br />Caused by: ptolemy.kernel.util.NameDuplicationException: Attempt to insert object named "plot" into container named ".configuration.directory.effigy2", which already contains an object with that name.<br /> at ptolemy.kernel.util.NamedList.append(NamedList.java:133)<br /> at ptolemy.kernel.CompositeEntity._addEntity(CompositeEntity.java:2080)<br /> at ptolemy.kernel.ComponentEntity.setContainer(ComponentEntity.java:518)<br /> at ptolemy.kernel.CompositeEntity.setContainer(CompositeEntity.java:1815)<br /> at ptolemy.actor.gui.Effigy.setContainer(Effigy.java:401)<br /> at ptolemy.kernel.ComponentEntity.<init>(ComponentEntity.java:106)<br /> at ptolemy.kernel.CompositeEntity.<init>(CompositeEntity.java:164)<br /> at ptolemy.actor.gui.Effigy.<init>(Effigy.java:133)<br /> at ptolemy.actor.gui.PlotEffigy.<init>(PlotEffigy.java:71)<br /> at ptolemy.actor.lib.gui.Plotter.initialize(Plotter.java:155)<br /> ... 2 more<br />Caused by: ptolemy.kernel.util.NameDuplicationException: Attempt to insert object named "plot" into container named ".configuration.directory.effigy2", which already contains an object with that name.<br /> at ptolemy.kernel.util.NamedList.append(NamedList.java:133)<br /> at ptolemy.kernel.CompositeEntity._addEntity(CompositeEntity.java:2080)<br /> at ptolemy.kernel.ComponentEntity.setContainer(ComponentEntity.java:518)<br /> at ptolemy.kernel.CompositeEntity.setContainer(CompositeEntity.java:1815)<br /> at ptolemy.actor.gui.Effigy.setContainer(Effigy.java:401)<br /> at ptolemy.kernel.ComponentEntity.<init>(ComponentEntity.java:106)<br /> at ptolemy.kernel.CompositeEntity.<init>(CompositeEntity.java:164)<br /> at ptolemy.actor.gui.Effigy.<init>(Effigy.java:133)<br /> at ptolemy.actor.gui.PlotEffigy.<init>(PlotEffigy.java:71)<br /> at ptolemy.actor.lib.gui.Plotter.initialize(Plotter.java:155)<br /> at ptolemy.actor.lib.gui.SequencePlotter.initialize(SequencePlotter.java:130)<br /> at ptolemy.actor.process.ProcessThread.run(ProcessThread.java:135)</p> Bug #5052 (Resolved): default file open directory of Kepler 2.0https://projects.ecoinformatics.org/ecoinfo/issues/50522010-06-17T00:38:51Zjianwu jianwujianwu@sdsc.edu
<p>If I remember correctly, the default directory for 'File-->Open File...' menu item is getting-started directory in Kepler 1.0. In Kepler 2.0 RC 5, the default directory is KeplerData and the getting-started directory is at KeplerData/workflows/module/outreach-2.0.0/demos/getting-started. It might be inconvenient to find this directory, especially for novice.</p>
<p>The discussion about it on IRC is as follows:<br /><jianwu> but why it has such a name, and it is not the default dir for 'file opening'<br /><jianwu> If I remember correctly, the file open will go to the getting-started directory by default.<br /><jianwu> for Keplre 1.0<br /><jianwu> In Kepler 2.0, this directory will take people a long time to find it.<br /><jianwu> what do you think, davidwelker?<br /><davidwelker> hmm<br /><davidwelker> I think we could change it<br /><davidwelker> who knows how to change it quickly?<br /><davidwelker> I agree this is less convenient. But I am not going to delay the release for it<br /><davidwelker> Oh wait...<br /><jianwu> not sure. may chad know the situation of Kepler 1.0 and how to do it?<br /><davidwelker> I don't know if we should use getting-started anymore as the defult<br /><davidwelker> what if outreach is patched and goes to outreach-2.0.1<br /><davidwelker> then you code might break<br /><davidwelker> that is sort of fragile<br /><davidwelker> yeah, lets just leave it<br /><davidwelker> I agree it is a little less convenient<br /><davidwelker> But it is fine for now<br /><davidwelker> I want to do a Kepler 2.1 in the not too distant future<br /><davidwelker> and we can fix stuff like this<br /><jianwu> hmm<br /><davidwelker> you feel strongly about this =)<br /><davidwelker> if you feel really strongly, we can move it up<br /><davidwelker> but the issue is tricky. We have to find where the default is set<br /><davidwelker> and then that person has to detect not outreach<br /><davidwelker> but the latest outreach<br /><davidwelker> and then go from there to demos/getting-started<br /><jianwu> you can set it acordding to property("outreach.workflowdir")<br /><davidwelker> I am not sure how exactly to do that in code right off the top of my head<br /><davidwelker> right<br /><jianwu> not like outreach-2.0.1 or outreach-2.0.0 <br /><davidwelker> you want to change it?<br /><davidwelker> I agree<br /><davidwelker> that is right<br /><davidwelker> I remember now<br /><davidwelker> derik made that<br /><davidwelker> so, if you want to put this out before 2.0, do you think you can hunt down where the default is and change it?<br /><davidwelker> if you don't have time, it would be fine to do it post 2.0<br /><davidwelker> I am not even sure that getting-started is where we want it to always open files<br /><davidwelker> that seems great for checking out demos<br /><jianwu> I can try. I'll fire a bug firstly about the issue.<br /><davidwelker> but what about their own work?<br /><davidwelker> If you are working on your own stuff, you want to open your own workflow<br /><davidwelker> so, should the default really be getting-started?<br /><jianwu> it's true. <br /><jianwu> so it should be configurable?<br /><davidwelker> yeah<br /><davidwelker> I think it should be configurable<br /><davidwelker> but, I think KeplerData would probably be good for now, actually<br /><jianwu> ok. good discussion. I'll fire an enhancement bug and see what other guys say about it.<br /><davidwelker> perfect<br /><derik_> just catching up here<br /><derik_> i think Open dialog opens whereever you last opened from<br /><davidwelker> yeah<br /><jianwu> yes<br /><davidwelker> it definitely does htat<br /><davidwelker> cause if I open one workflow from getting-started<br /><davidwelker> then it opens from there without me having to navigate to it<br /><derik_> ok...maybe that wasn't relevant to what you guys are discussing :)<br /><davidwelker> so, where should it start the first time?<br /><davidwelker> actually derik, it is relevant<br /><davidwelker> it puts it in context<br /><derik_> ok, hmm<br /><davidwelker> we are just talking about where it should open the first time<br /><davidwelker> so, getting-started is not a bad choice<br /><jianwu> I agree<br /><davidwelker> since if you are working on your own stuff, well, it is going to open near there anyway<br /><davidwelker> I will leave decision to whoever has an opinion. =) I can go either way<br /><derik_> yeah getting-started would be a good choice for first time open. it would be a bad choice for first time save (not sure if they're one and the same or not)<br /><davidwelker> good point derik_<br /><davidwelker> I don't think getting-started would be a good choice for first time save</p> Bug #5020 (Resolved): 2.0-RC3 installer: R Kepler Module is not optionalhttps://projects.ecoinformatics.org/ecoinfo/issues/50202010-05-20T17:28:08ZChristopher Brookscxh@eecs.berkeley.edu
<p>When installing Kepler-2.0-RC3, I'm presented with<br />two choices:<br />Kepler<br />R Kepler Module</p>
<p>Both are required, I can't uncheck "R Kepler Module",<br />so why bother with this window?</p> Bug #4711 (Resolved): Problem with L.H. actor tree: ActorName,karhttps://projects.ecoinformatics.org/ecoinfo/issues/47112010-01-29T16:41:18ZChristopher Brookscxh@eecs.berkeley.edu
<p>In a clean checkout under Windows, I'm now seeing the<br />kar files listed in the left hand actor tree.</p>
<p>I see<br />-Actors<br /> -ANOVA,kar<br /> ANOVA</p>
<p>I would expect to see<br />+Actors<br /> ANOVA</p>
<p>It would be nice to not display the ANOVA,kar nodes for kar files that<br />only contain actors.</p>