[Open-bio-l] Fwd: [Utilities-announce] NCBI Revised E-utility Usage Policy

[Peter]

On Thu, Mar 25, 2010 at 11:07 AM, Ewan Birney <birney at ebi.ac.uk> wrote:
>
> Oddly, I don't think you want to be registering BioPerl as a client
> with an email. Rather the Bioperl libraries should prevent a client
> programmer from using the functions without an email and program type
> entered. This forces the decision onto the client programmer.
>

I think for most cases, having a Bio* mailing list as a default
Entrez email address is pointless (we have almost no
control over how end users will call the Entrez functions, if
they will use the history or not, etc). The current behaviour
of defaulting to no email but raising a warning seems OK.

As Ewan says (and the NCBI earlier said they would
required), changing this to make the email mandatory is
also a sensible option.

There is a special case for running the Bio* unit tests, where
it might make sense to include the developer's mailing list (and
maybe set the tool to something like "BioPerl-unittests" rather
than just "BioPerl").

Peter