[Bioperl-pipeline] DataMonger

[kiran]

>
> No, absolutely I totally completely agree, overloading in SQL should be
> "flattened" in XML.
>
> What I was worried about is that a whole concept could change in XML,
> rather than just flattening/prettifying, e.g. datamonger not being another
> analysis but being attached to an analysis, that is a shift in semantics,
> not just avoiding overloading...

ok my understanding was that, we decided to run datamonger as an analysis
because of our implemenation (running it on nodes, taking case of
failures..) but it is actually conceptualised as being a data-preprocessor
to some particular runnable and so it should belong to that particular
analysis.. do you see it in a different way..

kiran