applications software metadata

[Nicholas Walton]

this looks exactly correct. the registry needs to allow a user or client
to discover 'VO enabled' applications (aka VO services) - which in the
simple cases are services which transform or generate data.

in turn the registry must present enough information to describe the
service, its interface and type, such that the service is able to be
called as part of a workflow, a stand-alone service, etc.

(a simple listing page of astronomy software packages with links to an ftp
download site for the source code is probably not quite what we want).

yours, nic

========================================================================
Dr N. A. Walton
(AstroGrid Project Scientist	       http://www.astrogrid.org)
(Euro-VO VOTC Project Scientist	       http://www.euro-vo.org)
Institute of Astronomy          Tel:   +44 1223 337503
University of Cambridge         Fax:   +44 1223 337523
Madingley Road                  WWW:   http://www.ast.cam.ac.uk/~naw
Cambridge, CB3 0HA              email: naw at ast.cam.ac.uk
========================================================================


On Fri, 27 Jan 2006, Roy Williams wrote:

> There are some FABULOUS ways to use the registry for applications, for
> example:
>
> -- Applications that can be used as part of a Astrogrid CEA workflow
> -- Applications that understand John Taylor's messaging protocol
> -- Applications that are remote SOAP services and have a WSDL file
> -- Applications that can handle VOEvent packets (my own favourite)
>
> In each case there is *formal* information in the registry record, so
> that a portal can use that information, eg. providing a menu of
> plug-and-play workflow components, or a clickable list of VOEvent
> publishers, or auto-generating a form for using a web service.
>
> We can think of a set of registries, each directed to a very specific
> kind of application that can be formally defined. That is MUCH more
> promising than a laundry list of "software that somebody somewhere
> might find useful for astronomy".
>
> Roy
>
> California Institute of Technology
> 626 395 3670
>