Kepler: Issueshttps://projects.ecoinformatics.org/ecoinfo/https://projects.ecoinformatics.org/ecoinfo/ecoinfo/favicon.ico?14691340362010-05-17T18:11:41ZEcoinformatics Redmine
Redmine Bug #5014 (Resolved): Add documentation to resources for the Workflow Run Manager modulehttps://projects.ecoinformatics.org/ecoinfo/issues/50142010-05-17T18:11:41Zdebi staggsstaggs@nceas.ucsb.edu
<p>Add resources > documentation into svn, and then place the workflow-run-manager documentation there. Documentation needs to be named the same as your module name, i.e. workflow-run-manager .pdf. This is so that the Modules Documentation for the tagging module can be found by that menu action in Kepler.</p> Bug #5013 (Resolved): Add documentation to resources for Tagging modulehttps://projects.ecoinformatics.org/ecoinfo/issues/50132010-05-17T18:10:18Zdebi staggsstaggs@nceas.ucsb.edu
<p>Add resources > documentation into svn, and then place the tagging documentation there. Documentation needs to be named the same as your module name, i.e. tagging .pdf. This is so that the Modules Documentation for the tagging module can be found by that menu action in Kepler.</p> Bug #4977 (Resolved): Can't associate multiple sets of reporting artifacts in one kar with differ...https://projects.ecoinformatics.org/ecoinfo/issues/49772010-05-05T00:00:07ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Reporting artifacts from imported runs are no longer being put into provenance since reporting attributes are no longer stored in the WorkflowRun. I'll change the code to instead look for and insert all items found in the kar manifest that use handlers ReportLayoutKAREntryHandler or ReportInstanceKAREntryHandler.</p>
<p>Since reporting items dependOn the workflow and not the workflowRun, this is another tripping up point for multiple runs in one kar -- how will we know which reporting artifacts to associate with which run? I think putting multiple runs in one kar will require reporting artifacts dependOn the execution lsid.</p> Bug #4800 (Resolved): CacheManager can not insert WorkflowRun into cachehttps://projects.ecoinformatics.org/ecoinfo/issues/48002010-02-17T20:41:12ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>When you export a run, CacheManager's insertObject cannot insert WorkflowRun entries into cache. This is because the call to serializeObjectInFile fails, caused by:</p>
<p>Caused by: java.io.NotSerializableException: java.lang.ref.WeakReference<br /> at java.io.ObjectOutputStream.writeObject0(ObjectOutputStream.java:1081)<br /> at java.io.ObjectOutputStream.writeObject(ObjectOutputStream.java:302)</p> Bug #4790 (Resolved): cmd-click in wrp does not create relationhttps://projects.ecoinformatics.org/ecoinfo/issues/47902010-02-11T22:43:52ZDaniel Crawldanielcrawl@gmail.com
<p>On a mac, doing a cmd-click in the canvas creates a relation (diamond icon) both in the kepler suite, and kepler 1.0, but not in wrp.</p>
<p>Also, the tooltip for the relation toolbar button says "Control-click", which is wrong for mac.</p> Bug #4600 (Resolved): Changing workflow & running causes "execution history not found" message to...https://projects.ecoinformatics.org/ecoinfo/issues/46002009-12-04T01:21:03ZOliver Soongsoong@nceas.ucsb.edu
<p>This uses the example workflow in bug 4598.</p>
<p>Run the workflow. Change "Figure Number" to 12 and run again. Now view the report. The "No execution history found" message should appear. The displayed report is the old one.</p>
<p>This may be related to bug 4599.</p> Bug #4599 (Resolved): changing Expression and running problematic workflow generates blank reporthttps://projects.ecoinformatics.org/ecoinfo/issues/45992009-12-04T00:54:20ZOliver Soongsoong@nceas.ucsb.edu
<p>This uses the same workflow in Bug 4598. Open the workflow and change the "Figure Number" Expression from 1 to 12. Run the workflow and view the output. It should be a single blank page with a time/date stamp at the bottom.</p> Bug #4574 (Resolved): Able to delete workflow tags when on a runhttps://projects.ecoinformatics.org/ecoinfo/issues/45742009-11-24T00:11:55ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>I'm able to delete a workflow tag when I shouldn't be to, that is after that workflow has been run. This used to work right; it wasn't possible.</p>
<p>1) open kepler<br />2) tag blank workflow "workflowTag" <br />3) execute, dismiss error<br />4) click on run, tag run "runTag1" <br />5) at this point, when context is still run, you can delete workflowTag</p>
<p>If you Export Archive this run, the workflowRun moml doesn't have the workflowTag, but the workflowTag is in the workflow moml, and undesired and inconsistent state.</p> Bug #4539 (Resolved): Applying a tag can be slowhttps://projects.ecoinformatics.org/ecoinfo/issues/45392009-11-10T02:45:37ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>I've tracked down (one of?) the reason(s) for applying a tag being slow -- starting at the Workflow Run Manager a call is eventually made to SQLQueryV8's getTagClassesForExecutionLSID, and the call therein to NamedOntClass.createNamedOntClassFromURIs(tagURI, ontologyURI) is the expensive part, it takes about 260ms. Sean do you know if that method can easily be sped up? An alternative might be to only (look up and) set a WorkflowRun's tags member variable right before it's really needed, but this feels a little less clean.</p> Bug #4528 (Resolved): BREAK_BEFORE property for report sections does not workhttps://projects.ecoinformatics.org/ecoinfo/issues/45282009-11-04T06:14:09ZOliver Soongsoong@nceas.ucsb.edu
<p>I'm not sure if this is broken or just hasn't been implemented, but if I create a trivial workflow and create a trivial report (text, section, text) and set the BREAK_BEFORE property on the section to "true", there doesn't seem to actually be a break before the section.</p> Bug #4510 (Resolved): can't add local repository if its name begins with the name of another loca...https://projects.ecoinformatics.org/ecoinfo/issues/45102009-10-28T19:24:57ZOliver Soongsoong@nceas.ucsb.edu
<p>If I've added C:\Documents and Settings\soong\Desktop\kruger as a local repository, I can't add C:\Documents and Settings\soong\Desktop\kruger2. It complains that one is a subdirectory of the other, even though they are not.</p> Bug #4447 (Resolved): Change Icon for Dynamic Report Itemshttps://projects.ecoinformatics.org/ecoinfo/issues/44472009-10-12T22:13:59Zdebi staggsstaggs@nceas.ucsb.edu
<p>Make item for Dynamic Report Items more representative of an actor / worflow component icon.</p> Bug #4410 (Resolved): Cannot open KARs from arbitrary location on diskhttps://projects.ecoinformatics.org/ecoinfo/issues/44102009-09-24T18:12:15ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>You cannot open a KAR from an arbitrary location on disk, it must reside in a local repository.</p>
<p>One work-around is to create a dialog that prompts the user to move such a KAR into a local repo when they try to open it.</p>
<p>If we're going to keep this restriction in place, I would also think the Save Archive dialog shouldn't allow, or at least make difficult, the saving of KARs in locations besides local repositories.</p> Bug #4247 (Resolved): Application menu items sometimes disappear on OS Xhttps://projects.ecoinformatics.org/ecoinfo/issues/42472009-07-14T19:49:25ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>I'm not sure the cause of this yet, but occasionally lately I've been finding myself with a kepler window open, but only the Kepler menu appears in the (system wide top) menu bar in os X -- i.e. "Kepler" shows, but File, Edit, View, Workflow, Tools, etc are all missing.</p> Bug #3148 (Resolved): Add authentication error reportinghttps://projects.ecoinformatics.org/ecoinfo/issues/31482008-02-14T19:06:12Zben leinfelderleinfelder@nceas.ucsb.edu
<p>When using the AuthenticationManager with the ecogrid services there is currently no feedback given to the user when errors occur (bad password, wrong username, service unavailable...etc).<br />Should use the MessageHandler's alert mechanism so that it can function without a GUI.</p>
<p>Note that when using the AuthenticatedQueryService, there is no requirement that the authentication credential be <em>valid</em>. Therefore an error that occurs during authentication should not prevent the authenticated query from continuing (equivalent to performing a public query).</p>
<p>Also: we may want to consider adjusting the login prompt behavior so that the prompt will reappear after an unsuccessful login attempt. Only after successfully authenticating (or canceling the action) would the prompt not reappear. Currently you are only given one chance to login - there is no re-prompting.</p>