OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

provision message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: RE: [provision] Attribute Grouping Was: One or many data models?


I understand your concerns Jeff.

And I agree that putting this sort of information at the data layer wouldn't
be the right thing to do architecturally. But, IMHO, it is a limitation of
the real-world use of SPML as a standard for defining the interoperability
of provisionable resources. As we all know, at some point, some subset of
this information will get presented to the user, and I'm sure we've all gone
down that road of solving this problem in our own ways. To continue to have
vendor-specific definitions of this information is robbing the user of the
value of accessing a "pool" of resources.

So, while the SPML definition itself is the wrong place for it, perhaps
there needs to be a sister specification for the presentation layer
metadata? Food for thought?

Until then, I'll focus my interests back on the matter at hand ... schemas
:-)

-Sandy


-----Original Message-----
From: Jeff Larson [mailto:Jeff.Larson@waveset.com]
Sent: Wednesday, March 03, 2004 1:45 PM
To: Sandy Walsh; Jeff Bohren; provision@lists.oasis-open.org
Subject: RE: [provision] Attribute Grouping Was: One or many data models?



I would strongly advise not putting presentation-oriented information
in the schema language.  Instead, if you need it, put it into the
model that is described by the schema.  In our experience, display
preferences are a rather arbitrary thing and are often different
depending on who is interacting with the system.

It's also a lot more complicated than just grouping attributes.  There
is labeling, value transformation, selective hiding, section header titles,
localization, widget selection (text box vs. radio button, etc.).
Speaking for Sun, we would never use an attribute grouping mechanism
defined in the SPML schema.  Instead we would request some sort of
presentation formatting object that is independent of the object
being edited.

SPML really shouldn't be in the business of defining a page
layout description language, and anything we would do short
of that wouldn't solve enough of the problem to be useful.

Jeff

> -----Original Message-----
> From: Sandy Walsh [mailto:sandy.walsh@abridean.com] 
> Sent: Wednesday, March 03, 2004 11:19 AM
> To: 'Jeff Bohren'; provision@lists.oasis-open.org
> Subject: [provision] Attribute Grouping Was: One or many data models?
> 
> 
> Good question Jeff.
> 
> I know for the xs:complex definition you can specify an 
> xs:group for the classifications of elements, so that should 
> work (think of it as a class with a series of inner-struct's).
> 
> However, the real challenge, and one that applies to a flat 
> set of attributes as well as grouped attributes, is the 
> ordering of the attributes. How to specify that the 
> presentation order should be First Name, Middle Initial and 
> Last Name, as opposed to the order in which the attributes 
> were parsed. Also, since XML-Schema is very deterministic, 
> you can specify the <attribute> requirement, but not so much 
> 'permit<attribute id="foo"> but not <attribute id="blah"> '
> 
> So, yeah, it may have to live outside of the raw schema 
> definition. Unless one of the other schema definitions is 
> more liberal.
> 
> -Sandy
> 
> 
> -----Original Message-----
> From: Jeff Bohren [mailto:jbohren@opennetwork.com]
> Sent: Wednesday, March 03, 2004 12:16 PM
> To: provision@lists.oasis-open.org
> Subject: RE: [provision] One or many data models?
> 
> 
> Sandy,
> 
> The attribute grouping idea sounds very interesting. Along 
> the same lines the schema could provide locale specific 
> attribute names as well. Would you see that as useful?
> 
> If this was added to SPML 2.0, wouldn't that necessitate the 
> need for a custom schema notation? XSD could be used to 
> define the data, but I'm not sure if it could feasibly be 
> used to provide presentation information.
> 
> Jeff Bohren
> Product Architect
> OpenNetwork Technologies, Inc
>  
> Try the industry's only 100% .NET-enabled identity management 
> software. Download your free copy of Universal IdP Standard 
> Edition today. Go to www.opennetwork.com/eval.
>  
> 
> 
> -----Original Message-----
> From: Sandy Walsh [mailto:sandy.walsh@abridean.com] 
> Sent: Wednesday, March 03, 2004 10:58 AM
> To: provision@lists.oasis-open.org
> Subject: RE: [provision] One or many data models?
> 
> 
> Hey all!
> 
> Good discussion! I've finally come up for air from our last 
> release to start getting involved :-) hi!
> 
> While there are good arguments being presented for keeping 
> the SPML 1.0 schema in 2.0, we have to ask ourselves if we 
> want to carry this baggage forward forever. I agree with the 
> general sentiment of dropping the 1.0 schema data.
> 
> Schema is a well-solved problem via XML-Schema (and others) 
> and, as a group, we should be focused on provisioning 
> use-cases and not yet another schema language. For that 
> reason, I think that XML-Schema (or some pluggable 
> definition) should be our strategy going forward. If you look 
> at the bulk of the SPML 1.0 specification, more than 50% of 
> it is related to schema definition. By refactoring this out 
> of the spec, SPML would become much lighter and therefore 
> gain greater acceptance by being easier to adopt. 
> 
> As a side note, there are other schema-related issues we 
> should consider
> *around* provisioning such as attribute grouping. Attribute 
> grouping would provide hints to the presentation layer on how 
> to organize attributes. So, if you have hundreds of 
> attributes that may be available in the presentation layer, 
> the schema can provide hints as to how to logically group 
> them. Personally, I'd be more interested in seeing that in 
> the SPML spec. 
> 
> Just a couple of thoughts. 
> 
> Cheers!
> 
> Sandy Walsh
> Abridean, Inc.
> 
> 
> -----Original Message-----
> From: Darran Rolls [mailto:Darran.Rolls@waveset.com]
> Sent: Tuesday, March 02, 2004 1:18 AM
> To: provision@lists.oasis-open.org
> Subject: [provision] One or many data models?
> 
> 
> As discussed, the committee has to decide on the data model 
> for the 2.0 specification.  On the one hand, as prototyped by 
> Jeff Bohren at the last F2F meeting, we can devise a solution 
> that keeps the 1.0 DSML data model by adding an extensible 
> schema "root" that allowed for its coexistence with a new 
> "XML Schema" data model.  The "cost" of this model is 
> increased complexity.  On the other hand we can take a single 
> data model solution as proposed by Gerry Woods at the F2F and 
> base 2.0 on a pure XML Schema data model at the "expense" of 
> 1.0/ 2.0 compatibility and backwards support for 1.0 in a 2.0 
> compliant service. Please consider this issue and ask 
> questions/state preferences now.  I propose we hold a ballot 
> on this issue around the next committee con-call 3/16/04. 
> Thanks Darran
> 
> 
> --
> Incoming mail is certified Virus Free.
> Checked by AVG Anti-Virus (http://www.grisoft.com).
> Version: 7.0.225 / Virus Database: 262.1.3 - Release Date: 3/3/2004
> 
> To unsubscribe from this mailing list (and be removed from 
> the roster of the OASIS TC), go to 
> http://www.oasis-open.org/apps/org/workgroup/provision/members
/leave_workgroup.php.

-- 
Incoming mail is certified Virus Free.
Checked by AVG Anti-Virus (http://www.grisoft.com).
Version: 7.0.225 / Virus Database: 262.1.3 - Release Date: 3/3/2004

-- 
Incoming mail is certified Virus Free.
Checked by AVG Anti-Virus (http://www.grisoft.com).
Version: 7.0.225 / Virus Database: 262.1.3 - Release Date: 3/3/2004
 

<<attachment: winmail.dat>>



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]