RE: Meeting Summary - 26 March 2007

Hi Kevin 

Since the lid seems to have been taken off Pandora's box again, and at risk of a collective groan from the DDWG members who thought they had nailed it down, here are some further thoughts.

One of the things we have been trying to do is to work within the framework of the DI Glossary, because inventing our own slightly different glossary would not seem to be warranted or desirable. Another thing we have been trying to do is to make sure that the fruits of our labour are not limited to the context that the DI Glossary pre-supposes (i.e. there is a richer set of meanings of device description, which we'd like to be useful for, than would be presupposed by limiting ourselves to that context). Those factors, together with the nature of the discussion of the list led us - imo - to a somewhat backwards definition.

So much for history.

I agree that instantiation is a good route to understanding, however, I (like Rotan) can't wholly buy into your proposed wording because as I see it we place equal priority on our specific charter objectives, which relate both to Web browsing, as well as relating to being extensible to other activities.

I do agree with you (and with Rotan) that elaboration of the nature of the types of value is distracting in the place in which it currently occurs. Though there were reasons for including this, which we need not revisit.

So how about this:

A device is something that is useful for carrying out some prescribed set of activities in some prescribed circumstances.

A context is a formal definition of the activities and the circumstances in which they are to be carried out.

A device description is a formal definition of named attributes and their permissible values in some prescribed context. The attributes chosen to form a device description are descriptive of features that are relevant to the context.

The permissible values for an attribute may take the form of lists, ranges or other patterns.

For the DDWG a device is as defined in the DI Delivery Context [def] but with the restriction that it is normally used away from fixed locations and is manufactured specifically to be portable and usable while being moved [from the charter].

Analogy:

A knife is a device that is useful for cutting. [DI definition of device]

A cooking knife is a device that is useful for cutting in the course of preparing food. [DDWG definition of device]

A table knife is a device that is useful for cutting prepared food in the course of eating it. [Someone else's definition of knife]

The device description for a cooking knife would include attributes describing, among other things, whether it is serrated or not, its length, whether the blade has a return from the handle making it suitable for chopping, the material from which the blade is composed, the material from which the handle is composed, the means of attachment of the blade to the handle, whether it is sharp ...

The device description for a table knife shares many of the properties of a cooking knife (because, after all, they are both knives). It's unlikely to contain attributes describing whether it's useful for chopping or not. The description of the serrations is likely to be different too, since table knives with serrations are usually used for cutting meat (steak knives) whereas cooking knives with large coarse serrations are used for cutting bread whereas those with small fine serrations are meant for cutting tomatoes.

Time for lunch.
Jo

> -----Original Message-----
> From: public-ddwg-request@w3.org [mailto:public-ddwg-request@w3.org] On
> Behalf Of Smith, Kevin, VF-Group
> Sent: 29 March 2007 11:12
> To: public-ddwg@w3.org
> Subject: RE: Meeting Summary - 26 March 2007
> 
> 
> 
> Whilst I think Jo's abstract definition is fine, there should be an
> 'instance' definition for each particular context to make it clearer for
> users of DDWG and other contexts.
> 
> If we replace the placeholders in the abstract definition with the
> explanatory text for DDWG, rather than supplementing the abstract
> definiton, we get something like:
> 
> "A device description is a formal definition, within the delivery
> context[1], of the named attributes and their permissible values which are
> applicable to devices[2]."
> 
> ...where [1] and [2] link to the DI Glossary. Granted, 'delivery context'
> may be too broad, but you see what I mean.
> 
> NB having read it a few times I think "[which may take the form of lists,
> ranges, or other patterns]" makes for disjointed reading and is too much
> detail, 'permissible' implies a formal constraint and should suffice.
> 
> Cheers
> Kevin
> 
> 
> -----Original Message-----
> From: public-ddwg-request@w3.org [mailto:public-ddwg-request@w3.org] On
> Behalf Of Jo Rabin
> Sent: 29 March 2007 10:49
> To: Christian Timmerer (ITEC); public-ddwg@w3.org
> Subject: RE: Meeting Summary - 26 March 2007
> 
> 
> Hi Christian
> 
> I completely agree that explanatory notes would be helpful, and I will aim
> to add some to the Wiki Definition [1].
> 
> You ask if the term "some context" was used deliberately and the answer is
> "yes". As Rotan points out, he included only part of the definition in his
> summary. The full definition, which as I say is at [1], actually says:
> 
> "A device description is a formal definition within some context of the
> named attributes and their permissible values [which may take the form of
> lists, ranges, or other patterns] which are applicable to entities of
> interest in that context."
> 
> "In general that entity is a device in the context of interest."
> 
> "In the DDWG that entity is something that can fulfil the role of device
> as defined in the DI Delivery Context."
> 
> The reason for putting it this way is that the meaning of "device" can be
> different according to the context you wish to discuss. As far as the DDWG
> is concerned the context (and hence the properties of interest) is
> delivering the Web to mobiles. So a device is something that is capable of
> accessing the Web, from our perspective.
> 
> However, we don't want to exclude the use of the same framework in
> different contexts. For example, you might want to describe the properties
> of mobile RRS Readers. We'd like you to use the same device description
> framework to do that, albeit that your meaning of device and the
> properties that are of interest are not exactly the same as our meaning of
> device and the properties that are of interest to us.
> 
> I agree that explanation is needed and as I say will add some text at [1]
> to elaborate. If you have a suggestion as to how the definition itself
> could be made clearer I'm sure the group would like to hear it.
> 
> Thanks for your comments.
> Jo
> 
> 
> [1] http://www.w3.org/2005/MWI/DDWG/wiki/DeviceDescriptionDefinition
> 
> > -----Original Message-----
> > From: public-ddwg-request@w3.org [mailto:public-ddwg-request@w3.org] On
> > Behalf Of Christian Timmerer (ITEC)
> > Sent: 29 March 2007 09:39
> > To: 'Luca Passani'; public-ddwg@w3.org
> > Cc: christian.timmerer@itec.uni-klu.ac.at
> > Subject: RE: Meeting Summary - 26 March 2007
> >
> >
> >
> > Dear Luca, all,
> >   as someone who's not "deep inside esoteric W3C lingo" I cannot make
> this
> > observation that the definition means nothing.
> >
> > My two cents are as follows:
> >  - However, the definition could be extended by notes/examples that
> helps
> > the reader to better understand the definition.
> >  - The wording "some context" in first part of the definition causes
> some
> > confusion to me because to me it means that this context needs to be
> > defined
> > by those who are adopting this definition. I'm wondering whether this
> > interpretation is correct/intentional.
> >
> > Thanks.
> > Best regards,
> >  -Christian
> >
> > :--
> > :- Dipl.-Ing. Dr. Christian Timmerer
> > :- Department of Information Technology (ITEC)
> > :- Klagenfurt University, Austria
> > :- http://research.timmerer.com
> > :----------------------------------------------------------
> >
> > >> Visit the IT Campus Carinthia
> > >> http://www.it-campus.at
> >
> >
> > > -----Original Message-----
> > > From: public-ddwg-request@w3.org [mailto:public-ddwg-request@w3.org]
> On
> > > Behalf Of Luca Passani
> > > Sent: Thursday, March 29, 2007 8:52 AM
> > > To: public-ddwg@w3.org
> > > Subject: RE: Meeting Summary - 26 March 2007
> > >
> > >
> > >
> > > "A device description is a formal definition within some context of
> the
> > > named attributes and their permissible values [which may take the form
> > > of
> > > lists, ranges, or other patterns] which are applicable to entities of
> > > interest in that context,"
> > >
> > > this is not english. It means nothing. It may mean something to those
> > > deep
> > > inside esoteric W3C lingo. Certainly not suitable for a blog as it is
> > > unless
> > > you want people to laugh at DD's work.
> > >
> > > Luca
> > >
> > > -----Original Message-----
> > > From: public-ddwg-request@w3.org [mailto:public-ddwg-request@w3.org]
> On
> > > Behalf Of Rotan Hanrahan
> > > Sent: Thursday, March 29, 2007 1:25 AM
> > > To: public-ddwg@w3.org
> > > Subject: Meeting Summary - 26 March 2007
> > >
> > >
> > > [Weekly conference call, 26 Mar 2007.] "Device Description" described!
> > > Initial ontology almost ready for publication. Reviewing OMA
> documents.
> > > Need
> > > public input to vocabulary. Details follow:
> > >
> > > [DD Defined] A text to describe what is meant by "Device Description"
> > > has
> > > been formally agreed by the group and will become part of the DD group
> > > terminology. The definition will require an update to the DI Glossary.
> > > The
> > > key part of the definition is this: "A device description is a formal
> > > definition within some context of the named attributes and their
> > > permissible
> > > values [which may take the form of
> > > lists, ranges, or other patterns] which are applicable to entities of
> > > interest in that context," and goes on to mention the role of 'device'
> > > and
> > > the scope of the DDWG with respect to this definition. The full text
> > > will be
> > > made public on the wiki.
> > >
> > > [Ontology Documents] A set of Protégé files capturing an initial
> > > ontology
> > > have been created by Rhys, together with a sample XHTML document to
> > > summarise the information in a human-readable form. Rotan will be
> > > looking
> > > into providing an automated visual representation, and the set of
> > > documents
> > > are expected to be made public soon. This is not the Vocabulary, but a
> > > framework in which the DDR Vocabulary can be defined.
> > >
> > > [OMA Liaison] In response to the recent OMA liaison, a formal
> > > acknowledgement will be posted on the public mailing list.
> Furthermore,
> > > two
> > > members of the group (Jo and Andrea) will be reviewing the OMA
> > > documents as
> > > requested in the liaison statement.
> > >
> > > [Vocabulary] It was noted during the meeting that the group could use
> > > some
> > > more public input, so expect members to mention the DDWG process in
> > > blogs
> > > and other public channels over the coming weeks.
> > >
> > > [New Actions] (ACTION-21) Rotan to Liaise with DI and successors ref
> > > this
> > > definition (of DD). (ACTION-22) Rotan to "Dot-ify" section 3 from the
> > > ontology document to make graphical version. (ACTION-23) Jo to work
> > > with
> > > Andrea to prepare a draft response to OMA Liaison - and review
> > > architecture
> > > document.
> > >
> > > [Attendees]
> > > Rodrigo Garcia Acevedo (CTIC)
> > > Pontus Carlsson (Drutt)
> > > Rotan Hanrahan (MobileAware)
> > > Martin Jones (Volantis)
> > > Cedric Kiss (W3C)
> > > Rhys Lewis (Volantis)
> > > Jo Rabin (dotMobi)
> > > Kevin Smith (Vodafone)
> > > Mike Smith (W3C)
> > > Andrea Trasatti (M:Metrics/WURFL)
> > >
> >
> >
> 
>

Received on Thursday, 29 March 2007 11:06:50 UTC