Kepler: Issueshttps://projects.ecoinformatics.org/ecoinfo/https://projects.ecoinformatics.org/ecoinfo/ecoinfo/favicon.ico?14691340362008-04-15T19:18:44ZEcoinformatics Redmine
Redmine Bug #3221 (New): offset new windowshttps://projects.ecoinformatics.org/ecoinfo/issues/32212008-04-15T19:18:44ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>New blank workflow windows overlap exactly. It would be nice if each new window was slightly offset from the last.</p> Bug #3215 (Resolved): Spaces in KEPLER path causes build problemshttps://projects.ecoinformatics.org/ecoinfo/issues/32152008-04-09T00:59:04ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Aaron Schultz reports - when trying to build under windows (in eclipse) with spaces in the -DKEPLER variable, Config.XML.saveDOM gives an error that lib/config.xml does not exist. Spaces are replaced with %20 and that's not working for windows. When -DKEPLER and -DPTII are changed to not have spaces kepler builds.</p> Bug #3203 (Resolved): Add dateTypehttps://projects.ecoinformatics.org/ecoinfo/issues/32032008-04-02T18:13:46ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Add a date type to Kepler; being able to pass date tokens between a variety of actors would be nice. My original motivation was to feed them to plot and other display actors.</p>
<p>Here a response from Christopher Brooks Sep 24,2007 on the original dateType/plotting topic:<br />--------<br />Hi Derik,<br />You could use the ptolemy.data.ObjectToken to encapsulate a Date.<br />See ptolemy.actor.lib.security.KeyToken for an example of a Token.</p>
<p>In the type hierarchy, the Date type would be convertable to String,<br />but probably nothing would be convertable to Date.</p>
<p>Adding date support to Ptplot would appear to be easy, but doing a<br />complete job would be very difficult.</p>
<p>The easy part is that one can easily specify ticks that consist of<br />a label and value. Thus, you could specify "January 1, 1970 00:00:00" <br />as 0 and then specify times in seconds since then.</p>
<p>See<br /><a class="external" href="http://ptolemy.eecs.berkeley.edu/java/ptplot5.6/ptolemy/plot/demo/Marks.htm">http://ptolemy.eecs.berkeley.edu/java/ptplot5.6/ptolemy/plot/demo/Marks.htm</a><br />for an example that uses ticks.</p>
<p>The hard part is: given a set of arbitrary Java Dates, label the axes<br />nicely, where nicely is defined as:<br /> - text does not overlap. <br /> - text displays meaningful values. <br /> For example, if we are just looking at 15 seconds, do we<br /> care about the day, month and year.</p>
<p>The code that labels the axes has lots of art in it. That code was<br />written a very long time ago - before many of the layout managers were<br />present.</p>
<p>There is also this entire question of the provenance of data. Perhaps<br />we would like to be able to attach metadata, such as the date, to<br />data points. The tricky part here would be to not make the Token<br />classes too large. Also, I'm not sure what you would do with the<br />metadata.</p>
<p>If you have data streams that have irregular time stamps, then you<br />might want to look at timed domains such Discrete Event. It is<br />possible to use SDF and have each iteration represent a tick of the<br />clock, but if you have data streams that have times associated with<br />them, then SDF is probably not what you want.</p>
<p>_Christopher</p>
<hr /> Bug #3161 (New): progress meter when downloading data with RBNBToKepler actorshttps://projects.ecoinformatics.org/ecoinfo/issues/31612008-02-23T00:21:57ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>When the RBNBToKepler actors are downloading data, some kind of progress meter should appear.</p> Bug #3160 (New): RBNBToKepler actors should work within SDF, PN and DDFhttps://projects.ecoinformatics.org/ecoinfo/issues/31602008-02-23T00:17:09ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>The RBNBToKepler actors need to work in a reasonable way under at least the following three domains: SDF, PN, DDF.</p> Bug #3159 (Resolved): RBNBToKepler datasource actors should be merged into one.https://projects.ecoinformatics.org/ecoinfo/issues/31592008-02-23T00:13:20ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>The 3 RBNBToKepler actors should be merged into one.</p> Bug #3158 (New): RBNBToKepler actors need to timeout more quicklyhttps://projects.ecoinformatics.org/ecoinfo/issues/31582008-02-23T00:08:14ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>RBNBToKepler actors need to timeout more quickly when an RBNB server cannot be reached. It currently takes too long for this to happen, and blocks Kepler in the process.</p> Bug #3157 (Resolved): RBNBToKepler actors should support all three sink modeshttps://projects.ecoinformatics.org/ecoinfo/issues/31572008-02-22T23:58:53ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>RBNBToKepler actors should support all three sink modes: Subscribe, Monitor and Request. Currently only Request is supported -- RBNBToKepler simulates Monitor mode with Requests for the newest data in an endless do-while loop, and RBNBToKepler2 and 3 make one Request for a user-specified timeslice of data.</p>
<p>1) Subscribe: The server pushes new data as it is received to the sink application's designated channel map. All data is sent to the application in a continuous stream regardless of whether the application can keep up.<br />2) Monitor: The server pushes new data as it is received to the sink application's designated channel map. The server skips ahead, dropping data if necessary, if the sink application gets too far behind in its processing.<br />3) Request: The sink application sends the server a channel map containing the name of the channel, timestamp, and duration of the requested data. The server responds by returning the channel map with the requested data filled in.</p> Bug #3156 (Resolved): RBNBToKepler actors should only download data for connected ports.https://projects.ecoinformatics.org/ecoinfo/issues/31562008-02-22T23:45:38ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>RBNBToKepler actors should only download channel data for output ports that are connected.</p>
<p>Currently a requestChannelNames input port exists on the actor that read in a string array, and the actor only downloads the specified channels. Hopefully this can be made redundant.</p> Bug #3155 (Resolved): RBNBToKepler actors should allow different output formatshttps://projects.ecoinformatics.org/ecoinfo/issues/31552008-02-22T23:39:14ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>RBNBToKepler actors need the ability to switch between at least two different styles of data output from their output ports (each of which is associated w/ an rbnb channel):<br />1) a record of arrays<br />2) an array of records</p> Bug #3143 (New): dataFrame_R cache problem under parallelshttps://projects.ecoinformatics.org/ecoinfo/issues/31432008-02-11T20:26:29ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>I'm currently getting the following error with the demos/R/dataFrame_R.xml workflow when trying to run it under Parallels:</p>
<p>Error in file(file, "r") : unable to open connection<br />In addition: Warning message:<br />cannot open file '////.PSF//.Home//.kepler//cache////cachedata//urn.lsid.localhost.7a976669.0.0', reason 'Invalid argument' <br />Execution halted</p>
<p>All other R demos work fine. I've tried deleting .kepler and restarting to no avail. Others have reported success under Parallels.</p> Bug #3140 (Resolved): RBNBToKepler2 actor output port generation problemhttps://projects.ecoinformatics.org/ecoinfo/issues/31402008-02-07T20:02:56ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>This actor is having problems dealing with output port generation when pointed at "niagara-dev.sdsc.edu:3333" (an sdsc test server). Below my discussion of what I think are the issues are (as posted to the rbnb-dev list <a class="external" href="https://lists.sdsc.edu/pipermail/rbnb-dev/2008/000298.html">https://lists.sdsc.edu/pipermail/rbnb-dev/2008/000298.html</a>):</p>
<p>I have two RBNB questions:<br />1) Does one need to do a Fetch to get channel types?<br />2) Why does a Fetch sometimes reduce the number of channels in a channelmap?</p>
<p>I hope the answer to 1) is No. Is there a way to get type without first doing a Fetch? Below are some snippets that illustrate the two issues.</p>
<p>I first set up a registrationMap:</p>
<p>------------------------------<br /> _map.Clear();<br /> _registrationMap.Clear();<br /> _sink.RequestRegistration(_registrationMap);<br /> _sink.Fetch(_blockTimeout,_registrationMap);<br /> _numChans = _registrationMap.NumberOfChannels();</p>
<pre><code>for (int i=0; i < _numChans; i++){<br /> chanType = _registrationMap.GetType(i);<br /> System.out.println("chanType of _registrationMap chan " +i + " is: " + chanType);<br /> _map.Add(_chanNames[i]);<br /> }<br />------------------------------</code></pre>
<p>Unfortunately the types printed are all 9, i.e. string....not true! (If I try to GetType on _map instead, after the _map.Add, I get NullPointerExceptions).</p>
<p>So what I've been doing after the above is the below code. After the following Request and Fetch I can use GetType to receive the variety of channeltypes I expect--strings, float64s, etc.<br />However, I was hoping I could get this 'metadata' without actually fetching data.</p>
<p>The below also illustrates issue 2).</p>
<p>------------------------------<br /> _numChans2 = _map.NumberOfChannels();<br /> System.out.println("num chans of _map: " + _numChans2);</p>
<pre><code>// just get one small chunk of data so we are able to check the data types<br /> _sink.Request(_map, 0, 1, "oldest");<br /> _numChans2 = _map.NumberOfChannels();<br /> System.out.println("num chans of _map: " + _numChans2);<br /> _sink.Fetch(_blockTimeout,_map);<br /> _numChans2 = _map.NumberOfChannels();<br /> System.out.println("num chans of _map: " + _numChans2);<br />------------------------------</code></pre>
<p>When I point at niagara-dev.sdsc.edu the stdouts print:</p>
<p>num chans of _map: 106<br />num chans of _map: 106<br />num chans of _map: 100</p>
<p>The Fetch seems to be changing the number of channels in _map! Does that make sense? Does Fetch throw-away / auto-disregard certain channels? On my server this does not happen, num chans does not change in those three stdouts. I also don't think it's that rbnb is throwing away channels with duration 0--I see more channels than 6 that have duration 0 on niagara-dev.</p>
<p>This is with v3.1b1 and v3.1b2.</p> Bug #3130 (Resolved): Add an accelerator (keyboard shortcut) for opening a new blank workflowhttps://projects.ecoinformatics.org/ecoinfo/issues/31302008-02-05T21:28:19ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>Add an accelerator for opening a new blank workflow canvas (File => New Workflow => Blank).</p>
<p>Unless 'n' is taken I will use it. e.g. apple-n will be the accelerator on a mac for this action.</p> Bug #2889 (Resolved): fileName in ConfigXML.java not getting sethttps://projects.ecoinformatics.org/ecoinfo/issues/28892007-07-12T00:03:22ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>ConfigXML.java has two constructors, but only one sets String fileName, the variable that points to Config.xml. This variable is used later in the SaveDOM method and must therefore be set if one wants to be able to write to the Config.xml file.</p> Bug #2872 (Resolved): md5 checksumshttps://projects.ecoinformatics.org/ecoinfo/issues/28722007-06-11T23:16:45ZDerik Barseghianbarseghian@nceas.ucsb.edu
<p>We should add md5 checksums to our releases.</p>