[MOBY-dev] [MOBY-l] Cleaning the registry

[Pieter Neerincx]

Hi,

On 06 Jan 2009, at 14:43, Pierre-Yves Chibon wrote:

> Andreas Groscurth wrote:
>> Hi,
>> I uploaded to html files showing the unsused elements:
>> http://bioinfo.mpiz-koeln.mpg.de/datatypes2Delete.html
>> http://bioinfo.mpiz-koeln.mpg.de/namespaces2Delete.html
>> The corrected number for the datatypes is 331 / 721 (45%).
>
> I think there is still a bug, I see:
> OSE_object	avilin05.avignon.inra.fr	2008-08-20
> and I know this object (I created it), it has a bunch of children  
> (OSE_author, OSE_experiment...) that are used by webservices :)
>
> We can not remove object that have children that are used.

Check, same here. I've registered a more generic object which I don't  
use directly in any service, but I do use a more specialized object  
that inherits from this one (ISA relationship).

For all the other object and namespace "orphans" registered by www.bioinformatics.nl 
: they are actually used, but you can't figure that out using the  
ontologies (yet). Long time ago I ran into scalability issues. After  
trying several different solutions to the problem I started to use  
"pass-by-reference". Hence when I expect large inputs or outputs I  
replace them with a URL object. This URL points to a file somewhere on  
a server, which contains the actual data in BioMoby format. If one day  
pass-by-reference will become standardised as part of the official  
BioMoby specs, I expect to be able to take advantage of it with only  
minor tweaks as all the data I send around is already BioMoby  
compliant :). Registering those objects and namespace made sense to me  
because:

1. It allows me to use BioMoby tools to explore the registry +  
structure of the objects when designing/maintaining my own services.
2. Others might recycle these objects preventing the registration of  
objects/namespace which are virtually the same, but just with a  
slightly different name creating yet another alias.
3. It prevents others from registering objects/namespace with the same  
name, but a for a different purpose or with a different meaning. That  
would require me to rename lots of objects/namespace once pass-by- 
reference makes it into the BioMoby specs. Hence it makes my current  
services more potentially forward compatible...

On a side note: Martin, any news on an official pass-by-reference RFC?  
I remember some time ago we had lengthy discussion on the mailinglist...

I do realise that right now a BioMoby "enduser" can not see the  
difference between depreciated orphan objects/namespaces catching dust  
and the ones I registered with the future of a standardised pass-by- 
reference machanism for BioMoby in mind...

Cheers,

Pi

> Thanks for the lists,
>
> Regards,
>
> Pierre
> _______________________________________________
> MOBY-dev mailing list
> MOBY-dev at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/moby-dev

-------------------------------------------------------------
Wageningen University and Research centre (WUR)
Laboratory of Bioinformatics
Transitorium (building 312) room 1034

Dreijenlaan 3
6703 HA Wageningen
The Netherlands

phone:  +31 (0)317-483 060
mobile: +31 (0)6-143 66 783
e-mail: pieter.neerincx at gmail.com
skype:  pieter.online
-------------------------------------------------------------