[kepler-users] Bi-directional pipe (equiv. to loop?)
Ilkay Altintas
altintas at sdsc.edu
Thu Dec 1 16:53:45 PST 2005
Short answers interspersed below...
On Dec 1, 2005, at 11:36 AM, Maurice Yarrow wrote:
> Hello Kepler users:
>
> This is probably a bunch of trivial questions, but...
>
> Are any of the following possible:
>
>
> (1) A fires B, then (2) B fires A:
>
> ------- (1) -------
> | |----------->| |
> | A | | B | (I would assume this is
> | |<-----------| | trivially possible)
> ------- (2) -------
This might be domain dependent, but it is possible for many domains.
>
>
> A and B communicate on same pipe (like socket communication)
>
> ------- -------
> | | | |
> | A |<---------->| B | (I would assume this is
> | | | | not possible)
> ------- -------
>
You can have inout ports. So this would be possible sing those. I've
never used an bidirectional port myself.
> I assume this immediately above not possible because pipes (paths)
> are unidirectional (?)
>
>
> And lastly, can A conditionally send to either B or C ?
>
> ------- -------
> | | (true) |
> | A |----------->| B |
> | | | |
> ------- -------
> |
> | (false) (I would assume that this
> | is also trivially possible,
> V and that if this were NOT
> ------- possible, then what would
> | | be the point of workflow
> | C | at all !)
> | |
> -------
>
This is possible.
-ilkay
>
> Maurice Yarrow
>
> _______________________________________________
> Kepler-users mailing list
> Kepler-users at ecoinformatics.org
> http://mercury.nceas.ucsb.edu/ecoinformatics/mailman/listinfo/kepler-
> users
More information about the Kepler-users
mailing list