Re: publishing FPWD issue (was Re: Thursday telcon & agenda)

At 03:55 AM 1/18/02 +0900, Olivier Thereaux wrote:
>On Thu, Jan 17, 2002, Lofton Henderson wrote:
> > Conceptually, yes.  Only question is how to manage inter-part links with
> > "family" approach.
>
>Maybe that helps (maybe not), but working groups who use the family
>approach (I'm thinking of CSS here, Dimitris might have something to say
>about DOM) tend to re-publish all their documents at the same time.

Somewhat stronger than "tend to", I think we would *must* re-publish all at 
once if we take the family approach.  Even if some parts are unchanged.  If 
we didn't do that, an unchanged part which was "latest" would be linking 
into an obsolete version of a changed part.

Actually, isn't this the same with the multi-part approach?  With 
multi-part, we have to put all parts in one new dated directory when we 
change and republish one part.  With family, we have to put all (N) parts 
into N dated directories when we change and republish one part.


>A rather naive (?) question to the editors : how much of this
>cross-linking will there be? If there is not plenty of it, then it may
>be painful, but the pain shall not last long.

I don't know.  Let me ask Kirill and Dimitris to comment also.  Here is the 
layout of the documents:

1
2a      <->     2b
3a      <->     3b
4a      <->     4b

(1=Intro; 2=Proc&Ops; 3=Specifications; 4=Technical)
(a=Guidelines; b=Examples&Techniques)

Right now, there are something like 30-40 (currently uncompleted) links in 
#1, Frm:Intro (mostly in Ch.4), which point at either the other guideline 
parts (#2a, #3a, #4a), or to anchors in those parts.  I suspect that this 
might be the heaviest linking on this axis, i.e., between different topical 
parts along the "guidelines" axis.

Similar to the style of WAI standards, there will be lots of links from the 
"Guidelines" documents to their supporting "Examples & Techniques" 
documents, i.e., from the "a" parts to the "b" parts.  In WCAG, for 
example, there is a link from every checkpoint (in "a" document) to the 
corresponding technique (in the "b" document).  I'm guessing there will be 
something like 30 checkpoints in #2a.

So ... is this plenty, or not?  I.e., how painful would the family approach be?

-Lofton.

*******************
Lofton Henderson
1919 Fourteenth St., #604
Boulder, CO   80302

Phone:  303-449-8728
Email:  lofton@rockynet.com
*******************

Received on Thursday, 17 January 2002 18:25:22 UTC