[MOBY-dev] bug report: jMoby Registry refers to default instead of accessing private

[Paul Gordon]

Eddie is indeed correct.  This is especially true if you consider 
applications like Seahawk, which can access multiple registries at once, 
but whose data objects are obviously registry-specific. While it would 
be possible to make data objects registry-less at creation time, this 
also means that they could not be validated (like the error you got) 
until service invocation.  As the libraries are used in interactive 
applications, I felt this was too late a point to tell the user they 
messed up in creating the data.


-Paul

Edward Kawas wrote:
> I guess it does seem redundant, but if you consider that CentralImpl is used
> to interact with a registry and the other class is utilized when invoking
> the services, then there is a separation of concerns there.
>
> If you do create a Registry object and pass it to MobyDataType, then you
> will solve your problems (as Paul mentioned).
>
> Let us know if you need help with anything that we mentioned here.
>
> Eddie
>
> -----Original Message-----
> From: moby-dev-bounces at lists.open-bio.org
> [mailto:moby-dev-bounces at lists.open-bio.org] On Behalf Of Richard Finkers
> Sent: August-10-09 7:55 AM
> To: Core developer announcements
> Subject: Re: [MOBY-dev] bug report: jMoby Registry refers to default instead
> of accessing private
>
> Hi Eddy,
>
> I am using the Java API and not MoSes. Correct, I am using the
> MobyComposite Object.
>
> I did not try to create a Registry object, instead, I am using the
> CentralImpl object for service discovery. Do you suggest that the
> registry information then will be lost for service invocation? I would
> naively expect that initialisation of the CentralImpl object also would
> prime the Registry object since it seems redundant to me to do it twice.
>
> Richard
>
>
>
>
>
> On Mon, 2009-08-10 at 06:58 -0700, Edward Kawas wrote:
>
>   
>> Hi Richard,
>>
>> Are you running into this problem using MoSeS? You say composite MOBY
>> object, so does that mean you are using MobyComposite?
>>
>> Did you try creating a Registry object that is primed with your registries
>> information?
>>
>> Eddie
>>
>> -----Original Message-----
>> From: moby-dev-bounces at lists.open-bio.org
>> [mailto:moby-dev-bounces at lists.open-bio.org] On Behalf Of Finkers, Richard
>> Sent: August-09-09 5:07 AM
>> To: moby-dev at lists.open-bio.org
>> Subject: [MOBY-dev] bug report: jMoby Registry refers to default instead
>>     
> of
>   
>> accessing private
>>
>> Hi All,
>>
>> I am working on implementing an own moby client (based on the simpleclient
>> example / Java Moby API) in the backbone of my database system. For this,
>>     
> we
>   
>> use a custom biomoby repository.
>>
>> All goes fine (service lookup / service call) however in the default
>> situation I run into the following error while processing the XML:
>>
>> Problem: org.biomoby.shared.MobyException: Attempted to build a composite
>> MOBY object from XML, but the tag (MeasurementSet) does not correspond to
>>     
> an
>   
>> existing data type in the registry (default).  Please check the spelling
>>     
> and
>   
>> capitalization of the XML tag, to match a registered data type.
>>
>> This is logic since the object MeasurementSet was not registered in the
>> default (Calgary) registry but in our private project repository.
>>
>> I was able to hack a work-around by changing the DEFAULT_ENDPOINT and
>> DEFAULT_NAMESPACE in the CentralImple.java class to the ENDPOINT and
>> NAMESPACE of my main (project) repository.
>>
>> I was not able to trace back the origin of the bug (yet). Can any of you
>> help me with this?
>>
>> Cheers,
>> Richard
>>
>>
>> _______________________________________________
>> MOBY-dev mailing list
>> MOBY-dev at lists.open-bio.org
>> http://lists.open-bio.org/mailman/listinfo/moby-dev
>>
>> _______________________________________________
>> MOBY-dev mailing list
>> MOBY-dev at lists.open-bio.org
>> http://lists.open-bio.org/mailman/listinfo/moby-dev
>>
>>     
>
> Dr. Richard Finkers
> Researcher Plant Breeding
> Wageningen UR Plant Breeding
> P.O. Box 16, 6700 AA, Wageningen, The Netherlands
> Wageningen Campus, Building 107, Droevendaalsesteeg 1, 6708 PB
> Wageningen, The Netherlands
> Tel. +31-317-484165  Fax +31-317-418094
> http://www.plantbreeding.wur.nl/
> https://www.eu-sol.wur.nl/
> https://cbsgdbase.wur.nl/
> http://www.disclaimer-uk.wur.nl/
> _______________________________________________
> MOBY-dev mailing list
> MOBY-dev at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/moby-dev
>
> _______________________________________________
> MOBY-dev mailing list
> MOBY-dev at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/moby-dev
>
>
>