Kepler: Issueshttps://projects.ecoinformatics.org/ecoinfo/https://projects.ecoinformatics.org/ecoinfo/ecoinfo/favicon.ico?14691340362014-11-02T20:02:55ZEcoinformatics Redmine
Redmine Bug #6614 (New): parameter editor background color different on Linuxhttps://projects.ecoinformatics.org/ecoinfo/issues/66142014-11-02T20:02:55ZDaniel Crawldanielcrawl@gmail.com
<p>In the actor parameter editor dialog, the background color is blue for values if the string mode is true. If the parameter has choices and string mode is true, the background should also be blue, but on Linux it is white. An example is the Array Peak Search actor's scale parameter. The color is white on Linux in Kepler and Vergil, but blue on Mac in Kepler and Vergil. I also confirmed this behavior has changed since Kepler 2.0.</p> Feature #6434 (New): add workflow execution time and other metadata to report designerhttps://projects.ecoinformatics.org/ecoinfo/issues/64342014-02-28T21:11:23ZDaniel Crawldanielcrawl@gmail.com
<p>It would be nice if the report designer had drag and drop items for metadata stored in provenance such as the execution timestamp, who ran the workflow, etc.</p> Bug #6413 (New): error when switching to DDF inside Case refinementhttps://projects.ecoinformatics.org/ecoinfo/issues/64132014-02-07T21:14:52ZDaniel Crawldanielcrawl@gmail.com
<p>I get an exception when switching from SDF to DDF inside a Case Refinement. It seems to be caused by having an unconnected input port.</p>
<p>To reproduce:</p>
<p>1. open $PTII/ptolemy/actor/lib/hoc/test/auto/Case1.xml<br />2. add a new input port to Case<br />3. run the model<br />4. open the Case actor and replace SDF in each Refinement with DDF<br />5. run the model</p>
<p>The error is:</p>
<p>java.lang.IllegalArgumentException: hasToken() requires a positive argument.<br /> at ptolemy.actor.Mailbox.hasToken(Mailbox.java:149)<br /> at ptolemy.actor.IOPort.hasToken(IOPort.java:2061)<br /> at ptolemy.domains.ddf.kernel.DDFDirector.prefire(DDFDirector.java:649)<br /> at ptolemy.actor.CompositeActor.prefire(CompositeActor.java:1665)<br /> at ptolemy.actor.lib.hoc.CaseDirector.prefire(CaseDirector.java:267)<br /> at ptolemy.actor.CompositeActor.prefire(CompositeActor.java:1665)<br /> at ptolemy.actor.CompositeActor.iterate(CompositeActor.java:1082)<br /> at ptolemy.actor.sched.StaticSchedulingDirector.fire(StaticSchedulingDirector.java:211)<br /> at ptolemy.domains.sdf.kernel.SDFDirector.fire(SDFDirector.java:492)<br /> at ptolemy.actor.CompositeActor.fire(CompositeActor.java:450)<br /> at ptolemy.actor.Manager.iterate(Manager.java:787)<br /> at ptolemy.actor.Manager.execute(Manager.java:352)<br /> at ptolemy.actor.Manager.run(Manager.java:1202)<br /> at ptolemy.actor.Manager$PtolemyRunThread.run(Manager.java:1760)</p> Bug #6102 (New): actor classes not saved properly in KAR fileshttps://projects.ecoinformatics.org/ecoinfo/issues/61022013-09-25T19:57:20ZMatt Jonesjones@nceas.ucsb.edu
<p>Kenneth Jones <a class="email" href="mailto:kennethjone@gmail.com">kennethjone@gmail.com</a> reported:<br />In Ptolemy II it is possible to use Vergil to convert a component into a class, save the class, and use it (instantiate or subclass it) in other workflows in the future. The Kepler GUI seems to have the ability to convert a component to a class and, in the workflow in which you created the class, create instances or subclasses of it, but I don't see a way to save the class, so it can be used in other workflows in the future (In Vergil there is a "save submodel only" checkbox in the File/Save dialog). Can classes be saved and reused in Kepler? If so, how?<br />------------</p>
<p>I verified this by creating a class and then attempting to save that class in a KAR file. The KAR file gets created, but does not contain the class. Attempting to open the saved KAR file produces an error in Kepler saying it can not me opened. The contents of the KAR file show it clearly is missing the class file, as all it contains is the manifest file:</p>
<p><code>$ jar tvf MyRampTestClass.kar <br /> 714 Wed Sep 25 11:41:54 AKDT 2013 META-INF/MANIFEST.MF<br /></code></p>
<p>To fix, once you have created a class in a workflow, you should be able to right click on it and select 'Save Archive (KAR)...', which allows you to save in a kar file of your choice in your 'My Workflows' directory. That should then show up in the tree on the left, and be able to be dragged to workflows for use.</p> Bug #5697 (New): Workflow Run Manager can fail to Force Open a run kar that Kepler canhttps://projects.ecoinformatics.org/ecoinfo/issues/56972012-08-24T19:41:25ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>After connecting the Workflow Run Manager (2.3.0) to the kepler library and using it to download the attached run kar, it will fail to Force Open it. Nothing seems to happen (I can't see the logs on Mtn Lion-- need to check for errors). Yet if you manually download the run kar from the web, and open it from Kepler's File-> Open, the workflow opens, though the run does not get imported into the WRM as it should.</p>
<p>Note we should delete kars like this from the non-dev 'official' library (so things are 'clean' for users), since it was made with an unreleased, unversioned module in the mix:</p>
<p>reporting-2.3;keplerio;ssh-2.1;workflow-run-manager-2.3;provenance-2.3;tagging-2.3;workflow-scheduler-gui-1.0;kepler-2.3;outreach-2.2;r-2.3;loader-2.2;actors-2.3;directors-2.0;opendap-2.0;dataturbine-2.2;ecogrid-2.3;authentication-gui-2.2;gui-2.3;module-manager-gui-2.3;authentication-2.3;repository-2.3;job-2.1;io-2.0;data-handling-2.1;sms-2.2;component-library-2.2;util-2.1;event-state-2.1;core-2.3;common-2.3;module-manager-2.3;configuration-manager-2.3;kepler-tasks-2.3;ptolemy-kepler-2.3</p> Feature #5666 (New): Build system needs a target to build all common or blessed non-core moduleshttps://projects.ecoinformatics.org/ecoinfo/issues/56662012-08-07T18:07:33ZChristopher Brookscxh@eecs.berkeley.edu
<p>There should be a way to easily build all the common or "blessed" <br />non-core modules.<br />For example, "ant change-to -Dsuite=kepler" does not check out <br />reporting or tagging.</p>
<p>We should have notion of "blessed" modules, which I would say at a minumum<br />must meet two requirements:<br />1) There is someone on kepler-dev who is actively maintaining the module<br />2) There is at least one automatic test that at least does a smoke test<br />to ensure that the module works. More tests are better, but "ant test-blessed" <br />or something should do something with each module.</p>
<p>Having blessed modules would help in refactoring the tree because we<br />would have a better idea about what will break.</p> Bug #5645 (New): kar uploading should need user account.https://projects.ecoinformatics.org/ecoinfo/issues/56452012-08-06T20:15:43Zjianwu jianwujianwu@sdsc.edu
<p>Currently, anonymous users can upload their kars to Kepler kar repository. They could bring malicious actors and be harmful if other users download them.</p>
<p>We should have user account management for kar upload so at least the uploads can be traced.</p> Bug #5581 (New): Make UI improvements to Components Downloadhttps://projects.ecoinformatics.org/ecoinfo/issues/55812012-03-27T21:44:54ZDerik Barseghianbarseghian@nceas.ucsb.edu
When you download a kar from a remote repository via kepler, you typically:<br />*Click Sources and configure Preferences->Components to check-mark a remoteRepository
<ul>
<li>Search Components for your term (As the very slow search occurs (bug#5276), the entire Kepler UI locks up and no busy cursor is shown)</li>
<li>Right-click on a result kar, and select Download.</li>
<li>Click cancel to clear the search results</li>
</ul>
<p>The kar is placed in your MyWorkflows dir, but if this is the first kar put there, MyWorkflows does not show up in the Components tree until you take some other action that 'rebuilds' the tree, e.g. clicking the Build button beneath Sources->Components, or sav</p>
To fix this bug:
<ul>
<li>During Components search, only lock the Components panel, and show a busy cursor.</li>
<li>Display the MyWorkflows dir after a first KAR is downloaded into it.</li>
</ul>
<p>In the same ballpark is bug#4953. Also, strangely if you go through the scenario above, and subsequently save a new workflow into MyWorkflows, MyWorkflows continues to not show up, even though if you just start kepler fresh, don't do a remote search, and save a first workflow into MyWorkflows it will.</p> Bug #5510 (New): On Ubuntu, Kepler can disappear from the dock while still runninghttps://projects.ecoinformatics.org/ecoinfo/issues/55102011-10-06T21:16:38ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Sometimes on Ubuntu 10.04, Kepler can be up and running, but the dock will show no Kepler icon. If you have many windows open and the kepler window is buried, this can lead you to believe Kepler's not running.</p> Bug #5170 (New): uploading a workflow requires a refreshhttps://projects.ecoinformatics.org/ecoinfo/issues/51702010-09-03T19:48:36ZDaniel Crawldanielcrawl@gmail.com
<p>When a workflow is uploaded to a remote repository, the component tree must be searched again before the workflow appears. It'd be nice if the workflow appeared when the upload completed.</p> Bug #5153 (New): Save As... closes all windows associated with a workflowhttps://projects.ecoinformatics.org/ecoinfo/issues/51532010-08-17T18:04:51ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>If you're working on a complex workflow, it's not uncommon to have multiple windows open for the one workflow -- windows for different contained composites. When you Save Archive, a new KAR is written, all windows for the workflow are closed, and the new KAR is opened. This process is slow, and also does not open all your windows in the state they were at before save, only the top level window is opened, requiring you navigate through your workflow re-opening everything as you had it.</p> Bug #5129 (In Progress): adding and removing configuration properties could be made easierhttps://projects.ecoinformatics.org/ecoinfo/issues/51292010-08-06T21:21:25ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Now that we've shipped 2.0, if a developer wants to add a new configuration property to a config file, they must first check if the user's config file has the property before accessing it, and add it if not. This is a minor hassle if your code tries to use this property in a number of places, each of which might be reached first during the user's kepler session...you have to do the check and possible add at each place. Instead, maybe the configuration manager could, during startup, run through all active modules config files and add properties if necessary to the user's copies in KeplerData. Handing removing properties would be harder -- if we remove a property, and then a user reverts to an older version of Kepler, this CM method would have to know to add the property back for that version.</p> Bug #4978 (In Progress): Create additional 'module' property for any menu entries that are insert...https://projects.ecoinformatics.org/ecoinfo/issues/49782010-05-05T00:21:51ZSean Riddleswriddle@gmail.com
<p>As of revision r24221 and r24222-3 on the branch, there's a way to avoid those errors that look like:</p>
<pre><code>[run] ERROR (org.kepler.gui.MenuMapper:getActionFor:271) Exception trying to create an Action for classname: &lt;org.kepler.tagging.TagManagerAction&gt;:<br /> [run] null (java.lang.ClassNotFoundException: org.kepler.tagging.TagManagerAction)</code></pre>
<p>Those come from modules inserting menu entry information into the persistent gui module configuration and then the user removing that module from their active configuration. The system still tries to create that entry, which references a class that is no longer available. If a property named 'module' is created as a peer to 'name' and 'value', then the menu entry will only be generated if a module by that name is in the active configuration. See org.kepler.module.tagging.Initialize from the tagging module for an example of how to do this without hardwiring the module name.</p>
<p>So in order to eliminate error messages like the above, these additional properties need to be created when the menu entry is added to the gui configuration.</p> Bug #4953 (In Progress): Remote Components category sticks around sometimeshttps://projects.ecoinformatics.org/ecoinfo/issues/49532010-04-21T00:51:03ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>If I search KeplerDevRepository for "run2", I currently get back one result: run2-mclaughlin_par,kar, inside the new category of Remote Components.<br />If I right-click on the kar and select Download, and then clear my search by, on the mac, hitting the little x button, or by hitting the Cancel button, the Remote Components category stills shows up in the tree. If you don't select download, the Remote Components category disappears, as I believe is intended.</p> Bug #4049 (New): Import actor that already exists appends strange numbering schemehttps://projects.ecoinformatics.org/ecoinfo/issues/40492009-05-01T22:37:33ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>When I import an actor that already exists, saying Yes to use the next available id, the new actor appends a numbering scheme to the name that I don't understand. After three consecutive imports of the <strong>same</strong> DataTurbine actor, my tree now contains:<br />DataTurbine<br />DataTurbine_2_<br />DataTurbine_2__3_<br />DataTurbine_2__3__4_</p>
<p>I'm assuming that's wrong and should just be one incrementing number.</p>