Re: [codelist-comment] The basic structure of the lists

["Anthony B. Coates (Miley Watts)" <abcoates@mileywatts.com>]

[This is a personal reply, not an official reply of the Code List  
Representation TC.]

Dear David,

	The accusation that the Code List Representation TC, or anyone in OASIS,  
is actively misleading people is a strong one, and I would ask you to  
consider how strongly you wish to stand behind this claim.  Indeed, the  
name of the "Code List Representation TC" was chosen very specifically; it  
is not the "Code List TC" because we are producing a representation  
(interchange) format, not definitive run-time data structures for code  
lists.  There is also nothing in the specification (to the best of my  
knowledge) that suggests that genericode is intended as a run-time code  
lookup format (which I'm guessing is what you meant when you wrote  
"production codelist format", although the term "production" can also have  
other implications).

	I do think that warnings at the top of each genericode document are an  
unnecessary step.  However, I have made a note that the TC should consider  
adding more text in the specification so that it is even clearer what  
genericode is specifically designed for, and what it is not intended for.

	Cheers, Tony.

On Fri, 18 May 2007 14:18:03 +0100, David RR Webber (XML)  
<david@drrw.info> wrote:

> Martin,
> A further issue here is what I previously STRONGLY warned would happen -  
> and I see that it is indeed happening.
> I was assured that the genericode format would be only a reference  
> semantic information model and that it WOULD NOT be promoted as a  
> runtime format for actual operational codelists.  It clearly is  
> ill-suited to actual operational use as both you Martin and Tony have  
> noted already.
> OASIS and the TC needs to take strong steps in EVERY announcement on  
> availability and releases that these genericode formats are NOT designed  
> to be used as operational codelists - only reference sets to the  
> semantics that can be used in operational codelists that implementers  
> should derive from them.
> As a way of example - the shipping port codes genericode is a HUGE  
> multi-megabyte file, even compressed as a ZIP.
> A production extract from that in an optimized format - as you Martin  
> allude to - and such as OASIS CAM uses - can produce fine-grained lookup  
> lists from that in the order of 10k to 100k in size - depending on  
> application use needs.
> I feel obliged to say this again - at all costs here we should stop  
> misleading people into thinking genericode is a production codelist  
> format - and re-state for them that they are solely reference semantic  
> lists of codes.
>
> I would ask that a large an prominent WARNING comment statement is  
> placed at the top of each genericode instance so that people understand  
> the intent.
> Thanks, DW

-- 
Anthony B. Coates
Senior Partner
Miley Watts LLP
Experts In Data
+44 (79) 0543 9026
Data standards participant: genericode, ISO 20022 (ISO 15022 XML),  
UN/CEFACT, MDDL, FpML, UBL.
http://www.mileywatts.com/