[DAS2] Re: Apollo and DAS/2 priorities

Ed Erwin ed_erwin at affymetrix.com
Tue Feb 7 20:46:01 UTC 2006

This is something we should discuss when we discuss the 'writeable' 
parts of the spec.  But in my opinion, 'writeable' and 'lockable' are 
two separate <CAPABILITY>'s.  I see no reason not to allow some 
implementers to develop simple servers that are writeable but don't 
implement a locking mechanism.  Large public servers may want locking, 
but I'd bet that a non-locking server would very rarely lead to 
problems, especially in small projects.

(If the server is non-locking, the client could add a little more logic 
to check that nothing has changed since the last retrieval before doing 
a commit.)

Andrew Dalke wrote:
>>> There is no more 'writeable' (that's, IMO) something to be decided
>>> as part of the writeback spec.  It might be that we have a
>> i have not made the change if this is an IMO.
> Okay.  There is no "writeable".  The writeability is determined
> by the <CAPABILITY> element.  If there is a CAPABILITY with
> a type == "locks" then the server is (potentially) writeable
> in the same way that "writeable='yes'" means that it's writeable.
> Anyone else have an O?
>                     Andrew
>                     dalke at dalkescientific.com
> _______________________________________________
> DAS2 mailing list
> DAS2 at portal.open-bio.org
> http://portal.open-bio.org/mailman/listinfo/das2

More information about the DAS2 mailing list