[dojo-contributors] dijit.Tree assumes datastore implements Identity API

Michael Smith savoyboy at gmail.com
Tue Aug 21 10:34:01 EDT 2007


While I wasn't involved in the initial decision to have a separate Identity
API, I was under the impression that this was done because some datastores
may have data that don't have IDs. Coincidentally, all of the existing
datastore implementations DO implement the Identity API. But is there any
case where a datastore may need to read data to which it can't apply
Identities?

~Michael


On 8/20/07, Brian Douglas Skinner <skinner at dojotoolkit.org> wrote:
>
> > http://trac.dojotoolkit.org/ticket/4238
> >
> > dijit.Tree is accessing internal structures of the store instead
> > of going through the proper ... accessors, as well as making
> > assumptions that all stores implement Identity, which isn't true.
>
> This may not be a good idea, but I wonder if for 1.0 we should think
> about requiring all datastores to implement the Identity API methods,
> and then include those methods in the Read API and get rid of the whole
> notion of a separate Identity API.
>
> Back a year ago, when we first decided to have separate APIs for Read
> and Identity, the thinking was that the Identity API might grow to be
> fairly complicated, to deal with situations like relational databases
> with compound key fields.  But as it stands now, the Identity API is
> just 3 methods, all of which should be pretty easy for any datastore to
> implement.
>
> :o) Brian
>
>
> _______________________________________________
> dojo-contributors mailing list
> dojo-contributors at dojotoolkit.org
> http://dojotoolkit.org/mailman/listinfo/dojo-contributors
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.dojotoolkit.org/pipermail/dojo-contributors/attachments/20070821/89070e3f/attachment.htm 


More information about the dojo-contributors mailing list