RE: comment

More reasons to avoid multilingual properties on monolingual documents:

There's always a big cost in making data an array from a scalar
unnecessarily.  Queries (in general) against this data will be
more expensive, and your example usage postulates using this data
for that purpose.

Secondly, the logical model is weaker.  If you only want to look
at the German text in a resource, you must examine each natural
language property separately to see what language it is, rather
than looking in one place.  Not only is this a query speed problem,
but it also allows the case where none of the property values matches
the language or charset of the resource's contents.

Thirdly, this is inconsistent usage with DAV:display-name, and I
think you need a compelling reason to introduce this type of
discrepancy.

> -----Original Message-----
> From: ietf-dav-versioning-request@w3.org
> [mailto:ietf-dav-versioning-request@w3.org]On Behalf Of Clemm, Geoff
> Sent: Wednesday, January 24, 2001 9:45 PM
> To: ietf-dav-versioning@w3.org
> Subject: RE: comment
>
>
> I don't believe that the multiple comment strings were intended
> for property content negotiation, but rather to allow you
> to put comments in multiple languages on a resource which does
> not have multiple language variants.  This could help support
> searching for a document in multiple languages, even if the
> document itself is in only one language.
>
> I agree with Greg's point that a version is unlikely to have
> comments in multiple languages, but this property is intended
> for an arbitrary resource, not just for version resources.
>
> The reason this property (and DAV:creator-displayname) are
> specified this way is that they can take natural language
> values, which is not the case for the versioning specific
> properties.
>
> So I guess I see some benefit, with minimal cost.  To make
> sure I haven't missed anything, is there any cost to this
> extra layer of XML, other than a few extra bytes?
>
> Cheers,
> Geoff
>
> -----Original Message-----
> From: Tim_Ellison@uk.ibm.com [mailto:Tim_Ellison@uk.ibm.com]
> Sent: Wednesday, January 24, 2001 9:38 AM
> To: ietf-dav-versioning@w3.org
> Subject: RE: comment
>
>
>
>
> I agree with Eric that this approach for internationalizing individual
> properties is flawed, and that the change shoud be reversed.
>
> Regards,
>
> Tim Ellison
> Java Technology Centre, MP146
> IBM UK Laboratory, Hursley Park, Winchester, UK.
> tel: +44 (0)1962 819872  internal: 249872  MOBx: 270452
>
>
> "Eric Sedlar" <Eric.Sedlar@oracle.com> on 2001-01-23 05:25:58 PM
>
> Please respond to "Eric Sedlar" <Eric.Sedlar@oracle.com>
>
> To:   Tim Ellison/UK/IBM@IBMGB, ietf-dav-versioning@w3.org
> cc:
> Subject:  RE: comment
>
>
>
>
> I've just been looking into how to handle content negotiation, and I
> haven't seen any implementation where content identified by a single
> URL can handle multiple languages without redirection.  In other words,
> if my file has Dutch content, and I'm a French user, I need to be sent
> to another URL with the French version of that content (usually by the
> language extension on the filename).  That other resource would have the
> property values (like DAV:comment) in the correct language, presumably.
>
> Therefore, I think there is no need to worry about internationalization
> of string properties of a resource--they're going to be localized.  This
> change should be reversed.
>
> --Eric
>
> > -----Original Message-----
> > From: ietf-dav-versioning-request@w3.org
> > [mailto:ietf-dav-versioning-request@w3.org]On Behalf Of
> > Tim_Ellison@uk.ibm.com
> > Sent: Tuesday, January 23, 2001 5:17 AM
> > To: ietf-dav-versioning@w3.org
> > Subject: DAV:comment
> >
> >
> >
> >
> >
> > >   23.1.1 DAV:comment
> > >     Why do we have an extra DAV:string element?
> > >     Why can you have any number of them (how would a client
> > >    choose which one to display)?
> > >
> > > This was a change requested by Yaron to support
> > > internationalization.  You can have the comment
> > > string in multiple languages this way.
> >
> > I'll defer to those who are much more HTTP/XML-versed than I am, but I
> > would imagine that this would be a matter of content negotiation between
> > the client and server.
> >
> > Tim
> >
> >
> >
>
>
>

Received on Thursday, 25 January 2001 01:39:30 UTC