[MOBY-l] service = method ?

[Mark Wilkinson]

Gudmundur Arni Thorisson wrote:
> The
> Unix-philosophy, piping the output of one program into the next, is 
> something that I generally hold with, and therefore I like the whole 
> MOBY philosophy as well.

That is definately the "vision":

input->transform->output/input->transform->output/input->transform...

A transformation (i.e. service) that generates an output object type 
that is only usable by another program in its own service suite is not 
really a problem, but should only be done if there is some need for 
end-user observation or filtering of these intermediate objects, 
otherwise I would consider it to be a single service and do the "pipe" 
on the server-end, rather than register two separate services.


> But one question though, to the MOBY-savvy developer: could the service 
> provider define a 'super'-service of some kind, in MOBY-speech, that 
> uses two or more of its own methods implicitly, so that the user does 
> not have to bother with the middle steps? Something like what Mark wrote 
> earlier:
> 
>     $x = $y->operation1->operation2->operation3

Absolutely yes!  And now you are at the point of controversy that I am 
really struggling with right now:  How do you sensibly describe the 
transformation that these services are undertaking??  Remember, we have 
to somehow fit the service type description into an ontology!  When your 
input and output objects differ by more than one "intuitive" 
transformation it becomes very difficult to define a controlled 
vocabulary term for that transformation, and even more difficult to know 
where to fit it into a service type hierarchy!

As Charlie Brown would say... blech!

The solution might be to require services to be more "atomic", where the 
service must fit into our pre-defined type ontology, but I think we lose 
some of the MOBY magic by doing that - the idea of getting tangential 
information from a query is one of the nicest aspects of the MOBY 
solution, and that can best be done by "unusual" service types...

This is really the biggest octopus squatting on my brain these days, and 
it has been there for months!

M




-- 
--------------------------------
"Speed is subsittute fo accurancy."
________________________________

Dr. Mark Wilkinson, RA Bioinformatics
National Research Council, Plant Biotechnology Institute
110 Gymnasium Place, Saskatoon, SK, Canada

phone : (306) 975 5279
pager : (306) 934 2322
mobile: markw_mobile at illuminae dot com