[MOBY-dev] Cleaning out the registry

[Mark Wilkinson]

Hi All, 

As you have read, Martin and I are discussing the problem of cleaning up
the registry.  Though MOBY was not intended to be curated per se, it
isThere are four kinds of "bad" services in the registry right now:

1)  Services that are registered using some form of the word "test" in
their service name, service type, object name, and or namespace.  

2)  Services that do not work at all - 404 errors, or other clear
indications that the service is simply dead.

3)  Services that return invalid data-types (according to the recent
revision of the Object ontology).

4)  Services that may function perfectly well, but have been registered
entirely incorrectly (e.g. require plain-text input, but have been
registered as accepting base Object a input).  There are a HUGE number
of these in there, and they are quite annoying...


Regarding (1) and (2):  The main public registry is not intended to be a
laboratory, nor a lavatory!  I intend to remove these nuisance services
without any warning unless someone strongly objects and has a good
reason for objecting...  even so, as the host of my instance of the
registry, I feel it is within my jurisdiction to do my best to keep it
clean and make it useful, so the objection will have to be VERY good...

Regarding (3): Martin and I discussed this on-list the other day, and I
think it is appropriate that we contact the service provider as we
discover these errors, and if the service provider doesn't fix the error
in a reasonable time, then we delete the service.  At some point this
will become automated, when we begin adding sample input and output data
to the Service Instance RDF (an RFC will be sent to the list in a few
weeks relating to this)

Regarding (4):  I am loathe to fix the registration parameters of other
people's service, even in cases where the correct registration
parameters are obvious, because I don't want to become a full-time
curator of the Registry. It is also important to not change them because
I don't know what tools that service provider may have that are relying
on that service signature (even if it is wrong). Moreover, it is
important that service providers see their errors, and understand why
they are errors, such that their future registrations are correct.  Of
course, tools like Dashboard are going to make this all a piece of cake,
but in the meantime, I will be contacting service providers about
services that are obviously registered incorrectly, and I encourage
others to do the same :-)  However, I will not touch these services.

Please send comments ASAP, since I am about to submit a manuscript where
the reviewers will be using the registry, and I want to get it cleaned-
up as quickly as possible!

Best wishes all!

Mark



-- 
--
...his last words were 'Hey guys!  Watch this!'
--
Mark Wilkinson
Asst. Professor
Dept. of Medical Genetics
University of British Columbia
PI in Bioinformatics
iCAPTURE Centre
St. Paul's Hospital
Rm. 166, 1081 Burrard St.
Vancouver, BC, V6Z 1Y6
tel: 604 682 2344 x62129
fax: 604 806 9274

"For most of this century we have viewed communications as a conduit, 
       a pipe between physical locations on the planet. 
What's happened now is that the conduit has become so big and interesting 
      that communication has become more than a conduit, 
       it has become a destination in its own right..."

                Paul Saffo - Director, Institute for the Future