Reluctantly...

[Patrick Durusau <patrick@durusau.net>]

Greetings!

I have very reluctantly concluded that Dennis is correct about the lack 
of top down architecture in the current draft.

Part of the problem and delay has been that I kept trying to force the 
text, as written, into some semblance of being an architecture.

What I kept missing was that all the asides, the examples, and even the 
test cases, had as an agenda to specify what was lacking in the prose.

For example, the discussion yesterday that turns out to depend upon a 
sub-type of an array, known as a column or row vector. Those terms are 
in the text and the examples (sort of) make it clear what is meant, but 
that isn't how standards are written.

We had a similar situation with YEARFRAC and eventually that resulted in 
answers that I think will benefit everyone who uses YEARFRAC, rather 
than just saying this is the way it is. If you recall, David Wheeler 
came up with expressions that accurately generate the results for the 
various options for YEARFRAC.

I think the overall structure is sound, so we won't have to delay the 
CD-1 draft. It will mean that additional data types, for instance, will 
be necessary to specify the rules that now appear as asides, mentions, 
examples, etc.

The up side is that we will be able to clearly say what the rules are in 
fact and not depend upon someone noticing a prior example that gave the 
rule needed for some particular operation. The down side is that we need 
to isolate all those places so we can extract the necessary rules and 
formalize them.

The more quickly those places are identified and the rules stated, the 
more quickly a top down architecture can be fitted onto chapter 5 and 
verified.

Hope everyone is having a great day!

Patrick


-- 
Patrick Durusau
patrick@durusau.net
Chair, V1 - US TAG to JTC 1/SC 34
Convener, JTC 1/SC 34/WG 3 (Topic Maps)
Editor, OpenDocument Format TC (OASIS), Project Editor ISO/IEC 26300
Co-Editor, ISO/IEC 13250-1, 13250-5 (Topic Maps)