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

[Chris Fields]

Yep, tend to agree.  And the client program would be ultimately responsible for whatever is entered in.

Of course, we'll want to register certain tools/emails anyway ('BioPerl' possibly along with a white-listed email to the mailing list), just to be on the safe side.  Don't want the hassle of dealing with someone else pretending to be one of the various Bio*.

chris

On Mar 24, 2010, at 2:08 PM, Ewan Birney wrote:

> 
> Sorry to perk up here, but I think the right thing is to throw an exception
> if it's called without a "tool" and "email" parameter. Of course, then a client
> programmer can abuse this, but they are forced to put something in there.
> 
> 
> On 24 Mar 2010, at 18:21, Chris Fields wrote:
> 
>> Neither the eutils notification nor the new eutils docs make this very clear.  For instance, from reading the documentation, one would only have to register the tool and email once an IP is blocked.  However, later on it is indicated that the values supplied must be registered with NCBI or they will be blocked, which (to me at least) reads as if they must be registered regardless.  Which is it?
>> 
>> Also, there is the bit about the tool and email belonging to the software developer or organization, not the end-user, likely for the reasons Hilmar mentions.  Does this mean each tool has one assigned email?  This would then mean we need to either set both and register them just in case, or leave both empty and warn the user.
>> 
>> We have a bit of time to work out the specifics, just hoping NCBI responds (one never knows with them).
>> 
>> chris
>> 
>> On Mar 24, 2010, at 11:24 AM, Andy Jenkinson wrote:
>> 
>>>> From my experience, if you set a default value for something and there is very little advantage to changing it, people will rarely bother to do so.
>>> 
>>> The library developer's email address is not very useful for NCBI, who I assume wish to use it to contact whoever is consuming their resources. Being able to contact the Bio* developer doesn't really allow them to do this. The Bio* mailing list would be an option because there is at least some chance the app developer will get the email, but on balance I think it'd be better to incentivise people to change it themselves.
>>> 
>>> So I would say: leave it blank and give a warning.
>>> 
>>> Cheers,
>>> Andy
>>> 
>>> On 24 Mar 2010, at 14:08, Peter wrote:
>>> 
>>>> Hi,
>>>> 
>>>> This is probably of interest to all the Bio* projects offering access
>>>> to the NCBI
>>>> Entrez utilities. See forwarded message below.
>>>> 
>>>> I *think* the new guidelines basically say that the email & tool parameters are
>>>> optional BUT if your IP address ever gets banned for excessive use you then
>>>> have to register an email & tool combination.
>>>> 
>>>> Regarding the email address, the NCBI say to use the email of the developer
>>>> (not the end user). However, they do not distinguish between the developers
>>>> of a library (like us), and the developers of an application or script using a
>>>> library (who may also be the end user).
>>>> 
>>>> Currently we (Biopython) and I think BioPerl ask developers using our libraries
>>>> to populate the email address themselves. I *think* this is still the
>>>> right action.
>>>> 
>>>> Peter
>>>> 
>>>> ---------- Forwarded message ----------
>>>> From:  <utilities-announce at ncbi.nlm.nih.gov>
>>>> Date: Wed, Mar 24, 2010 at 1:53 PM
>>>> Subject: [Utilities-announce] NCBI Revised E-utility Usage Policy
>>>> To: NLM/NCBI List utilities-announce <utilities-announce at ncbi.nlm.nih.gov>
>>>> 
>>>> 
>>>> New E-utility documentation now on the NCBI Bookshelf
>>>> 
>>>> The Entrez Programming Utilities (E-Utilities) Help documentation has
>>>> been added to the NCBI Bookshelf, and so is now fully integrated with
>>>> the Entrez search and retrieval system as a part of the Bookshelf
>>>> database. This help document has been divided into chapters for better
>>>> organization and includes several new sample Perl scripts. At present
>>>> this book covers the standard URL interface for the E-utilties;
>>>> material about the SOAP interface will be added soon and is still
>>>> available at the same URL:
>>>> http://eutils.ncbi.nlm.nih.gov/entrez/query/static/eutils_help.html.
>>>> 
>>>> 
>>>> 
>>>> Revised E-utility usage policy
>>>> 
>>>> In December, 2009 NCBI announced a change to the usage policy for the
>>>> E-utilities that would require all requests to contain non-null values
>>>> for both the &email and &tool parameters. After several consultations
>>>> with our users and developers, we have decided to revise this policy
>>>> change, and the revised policy is described in detail at the following
>>>> link:
>>>> 
>>>> http://www.ncbi.nlm.nih.gov/bookshelf/br.fcgi?book=helpeutils&part=chapter2#chapter2.Usage_Guidelines_and_Requiremen
>>>> 
>>>> Please let us know if you have any questions or concerns about this
>>>> policy change.
>>>> 
>>>> 
>>>> 
>>>> Thank you,
>>>> 
>>>> The E-Utilities Team
>>>> 
>>>> NIH/NLM/NCBI
>>>> 
>>>> eutilities at ncbi.nlm.nih.gov.
>>>> 
>>>> 
>>>> 
>>>> _______________________________________________
>>>> Utilities-announce mailing list
>>>> http://www.ncbi.nlm.nih.gov/mailman/listinfo/utilities-announce
>>>> <ATT00001.txt>_______________________________________________
>>>> Open-Bio-l mailing list
>>>> Open-Bio-l at lists.open-bio.org
>>>> http://lists.open-bio.org/mailman/listinfo/open-bio-l
>>> 
>>> 
>>> _______________________________________________
>>> Open-Bio-l mailing list
>>> Open-Bio-l at lists.open-bio.org
>>> http://lists.open-bio.org/mailman/listinfo/open-bio-l
>> 
>> 
>> _______________________________________________
>> Open-Bio-l mailing list
>> Open-Bio-l at lists.open-bio.org
>> http://lists.open-bio.org/mailman/listinfo/open-bio-l
>