(Menaka - moving this from private email to dev@taverna as we try to
keep everything public - hope you are OK with that.)


COMDEV-120 proposes to make a generalized version of the scufl2 and
wfdesc command line tools.

My idea was that this would be an easy way into Apache Taverna,
without requiring a full understanding of the whole architecture right
away.


The Executeworkflow command line could certainly also deserve some
updates (the -help is 5 screens long!) - specially to support some
kind of mvn-like monitoring and status output while running (e.g.
"Finished processor X, took 149s", but also some updates for Taverna 3
(which profile and/or nested workflow to execute)  - this would take
you into more challenging territory with OSGi and calling the Taverna
Workflow Engine - "into the beast" so to speak. :-)


See Taverna 3 architecture diagram here:

https://www.lucidchart.com/publicSegments/view/5450cf71-3a74-4dc9-b0d4-22200a00c2c4/image.png
https://www.lucidchart.com/documents/edit/450c64b8-5166-808e-9c0b-55ba0a0057d0

(Greyed out things are desired/planned)

(BTW - Christian - where could this fit on the website?)


If you are up to the challenge, perhaps we can flesh out some
suggestion on what to update in executeworkflow as a phase 2? If the
scufl/wfdesc tool turns out to be easy, then it might not be enough
work for the whole summer.


On 9 March 2015 at 14:53, Stian Soiland-Reyes <[email protected]> wrote:
> Hi!
>
> Thanks for showing interest in our project proposal.
>
>
> I think the idea for the Taverna Language Command Line tool was for it
> to be separate from the existing Execute Workflow command line tool
> (the one you have looked at) - although it would be interesting to
> also consider the new functionality to be part of the command line
> tool distribution (e.g. as a separate launcher).
>
> The Taverna Language module is meant to also be used independently
> outside Taverna, and therefore has no dependencies on the rest of
> Taverna. As such it is relatively easier to package and build, and
> also it does not require the use of the OSGi plugin system that the
> executeworkflow includes.
>
> If you are looking at taverna-comandline for how we did command line
> options there - be aware that it includes a bit of trickery to work
> around OSGi threads and classloaders in passing in the command line
> arguments through the plugin system -- this should not be needed in
> the proposed Taverna Language command line as it would run with the
> regular classloader (either with a lib/*.jar collection or as a single
> executable jar.
>
>
> Note The executeworkflow program might still be slightly in a flux as
> we have been changing some of the dependencies recently, so it has not
> yet stabilized after move to Apache.
>
>
>
> On 7 March 2015 at 15:48, Menaka Madushanka <[email protected]> wrote:
>> Hi,
>> I downloaded the taverna language from GIT repository and built it.
>> And I went through the references that were given in
>> https://issues.apache.org/jira/browse/COMDEV-120?filter=12330297
>>
>> So, to clarify more about the project I'd like to get some ideas about what
>> are the command line options that taverna has.
>>
>> Also I need to know whether there should be a separate command line
>> interface.
>>
>> Thank you very much.
>>
>> Cheers...!!!
>
>
>
> --
> Stian Soiland-Reyes
> Apache Taverna (incubating)
> http://orcid.org/0000-0001-9842-9718



-- 
Stian Soiland-Reyes
Apache Taverna (incubating), Apache Commons RDF (incubating)
http://orcid.org/0000-0001-9842-9718

Reply via email to