Hi Stian,
Thank you for your detailed helps, I will try finishing up my implementation
with the information.
One thing to confirm, does the streaming input you described make things
efficient?
My purpose is not to load everything on the memory at once for the
efficiency,
but to recieve elements of a input list one by one,
output corresponding data one by one.
This is due to the assumed input/output could be a very large amount of
data, but can be divided into smaller chunks.
Seems like the last input you showed, a list of all of the elements, will be
loaded to the memory at the same time?
Best regards,
-Yoshinobu
On Tue, Oct 13, 2009 at 5:32 PM, Stian Soiland-Reyes <
[email protected]> wrote:
> On Tue, Oct 13, 2009 at 08:25, Yoshinobu Kano <[email protected]>
> wrote:
>
> >>> ERROR 2009-10-06 12:40:51,396
> >>>
> (net.sf.taverna.t2.workflowmodel.processor.dispatch.AbstractErrorHandlerLayer:166)
> >>> - Could not forget [...@48d334]
>
> > Generally speaking, what does "could not forget" in the error message
> mean?
> > Since this error is generated by Taverna, I would like to have any clue
> to
> > solve it -- at least to know whether it is something critical or not.
> > No stack trace, or any other information was dumped.
>
> It has to do with cleaning up state in the dispatch stack after an
> execution has finished.
>
> For some reason this clean-up is executed twice, but as the call is
> already finished this should not affect anything.
>
> Taverna developers: We could have a look at how the platform-version
> handles state.
>
>
> > Yes, assuming that my activity recieves a streamed input iteratively,
> > how can I notice the end of the iteration, from inside my activity code?
>
> Then you could make your activity expect depth 1 at the activity input
> port, with a granular depth at 0. (1: list of values, 0: individual
> values).
>
> That means that you expect the full list, but you don't mind being fed
> item by item.
>
>
> That means that you should get something like:
>
> [0] value0
> [1] value1
> [2] value2
> [] {value0, value1, value2}
>
>
> So the 'end of the iteration' if you like would be when you receive
> the full list in the end.
>
> I'm not sure if we've done an activity that receives inputs in this
> way, but it should be possible.
>
>
> > I am sorry I am not really familar with maven related settings.
> > Where should I put the plugins.xml file in my repository (assuming that
> it
> > is the same format of the one which Taverna's local folder contains) --
> the
> > root directory of the repository?
> > Under http://www.mygrid.org.uk/maven/,
> > http://www.mygrid.org.uk/maven/repository/, etc. I could not find any
> such
> > file.
> > And how about the update detection, is it simply based on the version
> number
> > in the plugins.xml file?
>
> There are two different repositories, which could be the source of
> some confusion.
>
>
> The Maven repository is where Taverna is to find your JAR and POM
> files. While developing you might want to use
> file:///your/home/directory/.m2/repository as the URL. (Browse to that
> folder using something like Firefox to get the URL syntax right!)
>
>
> Taverna looks for plugin definitions in plugin sites. Each known
> pluginsite is decribed in $HOME/.taverna-2.1b2/conf/plugin-sites.xml -
> typically you would need to add your own plugin-site.
>
> All the plugin-site needs to contain is a file pluginlist.xml - see
> http://www.mygrid.org.uk/taverna/updates/2.1-beta-2/plugins/pluginlist.xml
> for an example. - as you see there are then references to individual
> plugin files in the same folder, like to
>
> http://www.mygrid.org.uk/taverna/updates/2.1-beta-2/plugins/spreadsheet-plugin-0.2.0.xml
>
> What you see in there is the definition of the plugin. The version
> number and name etc. here are just used by the plugin manager. If you
> list a plugin with a plugin version higher than Taverna has installed,
> Taverna will allow you do an upgrade of that plugin.
>
> Internally in the plugin file you need to list which Maven
> repositories to use to find the software components, and also which
> components to include into Taverna. Int he case of the spreadsheet
> plugins the components are stated as these Maven dependencies:
>
> <dependency>
> <groupId>net.sf.taverna.t2.activities</groupId>
> <artifactId>spreadsheet-import-activity</artifactId>
> <version>0.2</version>
> </dependency>
> <dependency>
> <groupId>net.sf.taverna.t2.ui-activities</groupId>
>
> <artifactId>spreadsheet-import-activity-ui</artifactId>
> <version>0.2</version>
> </dependency>
>
> Now when you do a new plugin version you will also need to deploy a
> new version of those dependencies. As you see in this plugin there are
> two independent software components - meaning that you need to make a
> new plugin version when you have changed the version of at least one
> of those dependencies - allowing for you to upgrade in this case
> spreadsheet-import-activity-ui without updating
> spreadsheet-import-activity
>
> Be careful to not redeploy Maven artifacts with the same version
> number. Both Maven and Raven in Taverna will trust the version number,
> so if it finds that it has blablabla:0.2 downloaded in its local
> repository, it will not redownload it from your repository.
>
>
> While developing you would probably keep the Maven components in
> SNAPSHOT versions, like 0.3-SNAPSHOT. If you are using the
> file:///home/stain/.m2/repository/ in your plugins file then Taverna
> will not copy them to its own repository folder, so a new 'mvn
> install' would overwrite the JAR files picked up by Taverna. This
> means that you would not need to change the version number locally.
> However, if you suspect Taverna is using an older version, inspect the
> repository/ folder inside .taverna-2.1b2.
>
>
> > A bug report, probably you already know;
> > after running any workflow, "Workflow running" is still shown with a
> > rounding icon on the left.
>
> Is this for when the workflow has no outputs? - I believe that would
> be http://www.mygrid.org.uk/dev/issues/browse/T2-597 which Alex has
> just fixed
>
> If not, could you possibly send us the workflow that don't finish..?
>
>
>
> --
> Stian Soiland-Reyes, myGrid team
> School of Computer Science
> The University of Manchester
>
--
Yoshinobu Kano (Given/Family)
[email protected]
Project Research Associate, the University of Tokyo / U-Compare Project Lead
http://www-tsujii.is.s.u-tokyo.ac.jp/ http://u-compare.org/kano/
------------------------------------------------------------------------------
Come build with us! The BlackBerry(R) Developer Conference in SF, CA
is the only developer event you need to attend this year. Jumpstart your
developing skills, take BlackBerry mobile applications to market and stay
ahead of the curve. Join us from November 9 - 12, 2009. Register now!
http://p.sf.net/sfu/devconference
_______________________________________________
taverna-hackers mailing list
[email protected]
Web site: http://www.taverna.org.uk
Mailing lists: http://www.taverna.org.uk/taverna-mailing-lists/
Developers Guide: http://www.mygrid.org.uk/tools/developer-information
