[Biojava-l] LSID

[Tom Oinn]

mark.schreiber at novartis.com wrote:
> The internal use was mine (I was just using it as a substitute for a 
> namespace). Maybe we should upgrade it to be compatable with IBM or 
> Taverna?
> 
> - Mark
> 
> 
> 
> 
> 
> Michael Heuer <heuermh at acm.org>
> Sent by: biojava-l-bounces at portal.open-bio.org
> 06/16/2005 03:27 AM
> 
>  
>         To:     Gary Bader <bader at cbio.mskcc.org>
>         cc:     biojava-l at open-bio.org, Mark Schreiber/GP/Novartis at PH
>         Subject:        Re: [Biojava-l] LSID
> 
> 
> 
> The biojava LSID and the IBM LSID are slightly different APIs, the IBM one
> the more complete of the two.  There also are/were LSID client
> implementations that I'm not very familiar with in taverna [0] and for
> whatever reason in an email client called Haystack [1].

Taverna and Haystack (not an email client!) both use the reference 
implementation, i.e. the IBM one. In theory though the implementation 
shouldn't be that important, it's a standard after all - I'm not sure 
how actively supported IBM's one is but we've been using it quite 
happily for ages now.

We use LSIDs in a slightly different manner to that originally intended, 
in that we're mostly using them to name transient entities such as 
workflow process instances although we do also name concrete data items.

Cheers,

Tom (Taverna lead)