RE: Amendment to Holdings Schema ................................ .... ..........J.nr. 331-4

Sorry to get to this a little late -- Here's what I think:

I think Joe clearly described that yes, it is possible to express multiple
levels for a bibliographic part, using the built-in tree-like recursive
definition.  

As for the problem you mentioned, Leif, that you "would need special Danish
profiling of Holdings to reflect when a BibPart-structure reflects several
issues and when it reflects a single issue", I think you would simply need
to check out if your bib-part has child parts or not (cf.
'numberOfChildBibParts').  A single issue would be a 'leaf' (i.e. no
children) while a part representing several issues would be a non-leaf.   

Therefore, I don't think the proposed changes are needed.

Best regards,
Rolande 

-----Original Message-----
From: Pieter Van Lierop
To: 'Leif Andresen'; 'www-zig@w3.org'
Cc: 'Johan Zeeman'; Majordomo danZIG (E-mail)
Sent: 10/11/2001 5:36 AM
Subject: RE: Amendment to Holdings Schema ................................
.... 	..........J.nr. 331-4

I have not a strong opinion about this, but it seems to me that Joe is
right. 
Almost all the fields in the BibPart are optional, so you don't have to
repeat them. So I do not think that it is "very complicated" and I don't
understand why this is a DanZIG profile issue.

Maybe we can ask Rolande what she thinks about this?

Pieter van Lierop

> -----Original Message-----
> From: Leif Andresen [mailto:LEA@bs.dk]
> Sent: jeudi 11 octobre 2001 11:26
> To: 'www-zig@w3.org'
> Cc: 'Johan Zeeman'; Majordomo danZIG (E-mail)
> Subject: Amendment to Holdings Schema
> .................................... ..........J.nr. 331-4
> 
> 
> To the ZIG-list
> 
> On ZIG meeting last week our Proposed Amendment to Holdings 
> Schema of 11th
> September 2001 was discussed. 
> 
> It was decided to give a month respite for comments before Maintenance
> Agency approve the proposal or not.
> 
> On the ZIG-list Joe Z. comment it 27 September 2001. He's 
> point was that
> BibPart is already recursively defined and that the reason 
> for not approving
> the proposal.
> 
> It is correct that BibPart is already recursively defined. We 
> have discussed
> the possibility to use this approach.
> We have realised some problems with this model. We will then 
> need special
> Danish profiling of Holdings to reflect when a 
> BibPart-structure reflect
> several issues - and when is reflect a single issue.
> 
> If a single issue have the attributes:
> Vol 5, year 2001, issue 15, month May, week 8
> 
> With BibPart recursively defined you can do it with 3 or 5 BibParts.
> Only Enumeration and Chronology will be different for the 3 
> or 5 BibParts -
> all other tags will reflect the same information.
> It will make it very complicated and a profiling question to handle.
> 
> Our proposal:
> 
> Make Enumeration and Chronology recursive.
> 
> One BibPart
> Enumeration with one childEnumeration for Vol 5 and issue 15
> Chronology with two childChronology for year 2001, month May 
> and week 8
> 
> Best regards
> 
> Leif Andresen    
> Secretary danZIG
> Danish National Library Authority
> 
> > -----Oprindelig meddelelse-----
> > Fra: Johan Zeeman [mailto:joe.zeeman@tlcdelivers.com]
> > Sendt: 27. september 2001 23:44
> > Til: Poul Henrik Jørgensen; Ray Denenberg (E-mail)
> > Cc: ZIG Listserver (E-mail); DanZIG list (E-mail)
> > Emne: Re: Revised Holdings XML Schema v7
> > 
> > 
> > I'm afraid I haven't looked at this proposal closely before, 
> > but now that I
> > have I need to say that I think it stems from a 
> > misunderstanding of the
> > Holdings Schema.
> > 
> > For those who haven't looked, the proposal says that "It is 
> > not possible to
> > express more than one level for a BibPart.  An example:  One 
> > BibPart is
> > volume 5, issue 2. You can only express volume 5 or issue 2 
> > with version
> > 1.1. And it is a normal requirement to have several levels to 
> > present for
> > one BibPart."
> > 
> > However, BibPart is recursively defined.  That is, one 
> > BibPart can contain
> > child BibParts.  The intention of this is that a BibPart 
> > describes itself at
> > its own level and contains its children at lower levels.  The 
> > example would,
> > therefore, be expressed as a BibPart for volume 5 containing 
> > a child BibPart
> > for issue 2.  Each of these parts may contain additional information
> > pertaining to itself.  It is thus not "a normal requirement 
> > to have several
> > levels to present for one BibPart".  The "normal requirement" 
> > is that a
> > BibPart at any level may have several child BibParts, each of 
> > which may have
> > children of their own.
> > 
> > The proposal suggests remedying the defect by adding two new 
> > elements to
> > BibPart:  childEnumeration and childChronology.  Since I 
> > don't believe there
> > is a defect, I don't think there is any need for a remedy.
> > 
> > J.
> > 
> > ----- Original Message -----
> > From: "Poul Henrik Jørgensen" <PHJ@Portia.dk>
> > To: "Ray Denenberg (E-mail)" <rden@loc.gov>
> > Cc: "ZIG Listserver (E-mail)" <www-zig@w3.org>; "DanZIG 
> list (E-mail)"
> > <danzig@list.bibnet.dk>
> > Sent: Wednesday, September 19, 2001 8:17 PM
> > Subject: Revised Holdings XML Schema v7
> > 
> > 
> > > To Ray Denenberg, LC
> > >
> > > I have now prepared a revised version of the Z39.50 
> > Holdings XML Schema
> > > (HoldingsSchema7.xsd), that reflects the current proposed 
> > changes to the
> > > Holdings Abstract Schema.
> > >
> > > The revised Holdings XML Schema has been upgraded to 
> comply with the
> > current
> > > W3C May 2001 Schema Language.
> > >
> > > Both the new version 7 XML Schema with corresponding annotated
> > documentation
> > > as well as the previous versions are available via the same 
> > URL, that is
> > > referenced from the Z39.50 Maintenance Agency Web:
> > > http://www.portia.dk/zholdings/
> > >
> > > Best regards,
> > > Poul Henrik Jørgensen, DBC
> > 
>

Received on Monday, 29 October 2001 22:19:48 UTC