RE: [ubl-lcsc] pre-draft model for appraisal and initial change log

["Sue Probert" <sue.probert@dial.pipex.com>]

Hi Stephen and Tim

 

I agree that we need the utmost clarity before any editing of the model starts to be done and logged. My recollection of the agreeements arrived at during last Tuesday's call do not seem to tally exactly with what you mention below particularly with reagrds to the spreadsheet formatting decisions and the TBG17 submission details.

 

Tim, you are referring to what the minutes say but I cannot find the minutes for last week's call. Did they get sent out to the list? I apologise if I am the only person who did not receive them - my ISP is in super-suspicious mode right now to protect from virus attack and some messages are not getting through.

 

It seems to me to be very critical that we strike the right balance between the working methodology that we employed to produce the schemas for the beta release and what is needed now for the 1.0 release. I do not believe that we have achieved the required balance yet although my understanding of the agreeements reached last Tuesday is that they did clarify the way forward to achieve this.

 

During the recent ATG meeting Michael, with some help from me, spent a great deal of time QAing the beta spreadsheets and preparing the stored beta data models in EDIFIX in preparation for trialling automatic schema production. The exported spreadsheet circulated late last week by Michael represents a clear starting point for our 1.0 model editing. As I understand it, EDIFIX can export this same data in the format that we used previously thereby enabling us to generate schemas at any time via the scripts for comparison and validation. The question remains which excel format should be used as the basis for the editing work itself. This seems to be the area which is still in question.

 

Please could you or Bill post the minutes at the portal and circulate them again. I need to study them carefully before responding to both of your recent emails which I will do as soon as possible.

 

regards

Sue

 

-----Original Message-----
From: Tim McGrath [mailto:tmcgrath@portcomm.com.au]
Sent: 01 February 2004 07:47
To: Stephen Green
Cc: ubl-lcsc@lists.oasis-open.org
Subject: Re: [ubl-lcsc] pre-draft model for appraisal and initial change log

i am not sure what is 
going on hear but i also understood we had agreed not to change our current 
format.  The minutes state...

> After long debate WRT changing the spreadsheets to 
>   match TBG specific format, it was decided that it did not matter because the 
>   dictionary name that is generated in the UBL spreadsheet is not used.  
>   The tools (EDIFIX) generate these fields based on the TBG 17 macros (which we 
>   agree should be definitive in function – not methodology).  It would be 
>   nice if our formula produced the same results as the TBG 17 as we work on the 
>   content so that we can see immediate impact on dictionary name, but its not 
>   critical.  Keeping the format of the model as we now have it permits an 
>   opportunity for backward compatibility with other 
> tools.

what i see here is a TBG17  style spreadsheet 
with UBL data attached.  This is the wrong way around.  I can see that 
the values in the TBG17 side have been taken from the UBL side, so it should not 
be too hard to move the TBG17 entries to the other side (or preferably into a 
separate spreadsheet).  perhaps Michael can explain the problem with doing 
it this way?  If it is too hard then i suggest we continue with our current 
1.0-Beta sheets (after fixing up the embedded spacing issues).  the TBG17 
BIEs have been submitted and so we dont need to reflect this in our 1.0 final 
models.  in fact, we shouldn't because our models will be changing slightly 
and the TBG17 submission wont - so the inclusion of this here is misleading and 
incorrect.

the spreadsheet that EDIFIX produces should be the same of was 
published as 1.0 Beta.  as a test they should be capable of being processed 
by the UBLish script.  that is our sign of compatibility.  we agreed 
not to change this format.

i am also nervous about this all-in-one 
spreadsheet.  this is not how we have worked in the past and i cannot see 
why it needs to be any different for 1.0 final release.

on tuesday we 
must decide which models stephen should be applying changes to.

PS i dont 
think the datatype columns was used by the script anyway.  as it now 
appears un-necessary for TBG17 we should just use it to denote the 1:1 mapping 
of Representation Term to DataType - i dont think we should drop it from our 
models.

Stephen Green wrote:

Tim, Mikkel, LCSC,

 

Here, attached is a spreadsheet I'll be taking as a starting point for my changes from 1.0 beta to 1.0 draft 1.

I've called this version draft 0 as it should be equivalent to beta.

 

Mikkel had made changes:

combining all models into one spreadsheet

combining TBG17 and UBL model formats into the one spreadsheet

but there a few changes I've made myself,

the spreadsheet worksheet name to UBL

adding formulas as agreed in the meeting

plus I thought I'd better reverse the removal of the property noun and possessive,

as I seem to remember being agreed in the meeting

then I've also moved things a little and removed columns I thought were redundant

at the end of the TBG17 area (to keep it in line just with TBG17's format), etc

 

If these changes are disputed and amended via the Tuesday call this week I'll easily

be able to add the content changes to whatever is agreed, but I needed

a sure starting point as far as I think we've agreed.

 

Apologies if this means you have to amend the tool Mikkel, but I think this is what was

agreed in the meeting and the changes I think will be small

- though the columns may be in different places with some removed.

 

I did leave out the removed DataType-related columns from the UBL side,

so apologies again to those who might try using UBLish in future in that the

script will of course need changes beside NDR-related ones (unless empty columns are added first

before generating - along with removal of TBG17 columns and some rows at the top perhaps).

 

Now I'll start making the model content changes.

 

I've attached, in advance, my proposed initial changes to the contents.

 

The next version I'll call UBL 1.0 draft 1

 

All the best

 

Steve

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/ubl-lcsc/members/leave_workgroup.php.

-- 
regards
tim mcgrath
phone: +618 93352228  
postal: po box 1289   fremantle    western australia 6160