Attn. Brian - Re: [MOBY-l] MOBY Duple? :-)

[Jason E. Stewart]

"Mark Wilkinson" <mwilkinson at gene.pbi.nrc.ca> writes:

> "Jason E. Stewart" wrote:
> 
> > But even so, don't you think that it's worth calling out these bits
> > explicitly:
> >
> >   <Object auth="bioinfo.pbi.nrc.ca" ns="ArabidopsisGID" id="At2332487"
> >           sec="">
> >
> > and since 'seq' is optional it would be:
> >
> >   <Object auth="bioinfo.pbi.nrc.ca" ns="ArabidopsisGID" id="At2332487">
> >
> > I still think this is completely compatible with the LSID proposal,
> > it's just that we pre-parse the information into its components.
> 
> I think we'll have to defer to Brian on this one.  My impression
> (from reading comments in the I3C meeting minutes such as "they have
> their MOBY triple and they don't care") is that the LSID is intended
> to be used exactly as written, and that deviation from that, such as
> what we had with our triple + authority, is not on.  If we are
> allowed to parse it into our own format, and still claim that it is
> an LSID, then we have been ~compatible with it right from the
> get-go.  My impression was that we were not.
> 
> Brian - your opinion?

I think that we should attempt to convey exactly why this is a proper
use of an LSID.

> >   <Object auth="bioinfo.pbi.nrc.ca" ns="Arabidopsis/2.0/FooID" id="At2332487">
> >   <Object auth="bioinfo.pbi.nrc.ca" ns="Arabidopsis/2.1/FooID" id="At2332487">
> 
> yeah, that's probably a good idea... though it is breaking away from
> the keep things simple until they become necessarily complex :-)

Somewhat, I just know that it's going to happen, and I find the Unix
path name syntax much more legible.

jas.