RE: [oiic-formation-discuss] What are we validating and why?

Dave,

 

We learned on the NIH work that we actually had to handle about a dozen "flavours" of PDF - depending what had generated it, how and on what platform.  

 

I see that would be the same thing here - staying high level and avoid tool dependencies.

 

Therefore - from the scripting perspective that is all "thrash" under the hood - for them to handle.  True you need lowlevel code to grok the nuances - but the script can simply say :

 

use_config("NIH_page_defaults")

contains_pages("Section 5", 10, max) 

 

- or some such logic.  Implementations need to then take account of the rendering "thrash" - where tools deviate from the defaults - and if that causes error/warning/fail - and again that's something you would want to have configurable.

 

Thanks, DW

-------- Original Message --------
Subject: Re: [oiic-formation-discuss] What are we validating and why?
From: "Dave Pawson" <>
Date: Sat, June 07, 2008 2:01 am
To: 

2008/6/6 David RR Webber (XML) <>:
> Bart,
>
> Actually rendering for our NIH needs is validated in the big picture at the
> style level.
>
> So you are not allowed to have text in font smaller than 10 pitch and line
> spacing less than 1.
>
> Also page count totals per section similarly. So its not what it "looks"
> like - but specific aspects of how it is formatted.

Thanks, thats far clearer for me. And verifiable!
Do you see this as ODF testing or a user group testing for interop?
If two products produce a different page count would you say one
was 'wrong' or had failed the test?
This is close to a page layout issue, not an ODF issue IMHO.
Perhaps within a tolerance? pagecount = X +-2%.

>
> Then within the content - specific section headings, sub-headings and so on.

Capable of producing such headings, or must contain same?
Former perhaps within ODF scope, latter a user group issue?

>
> This requires tools that can "reach inside" the document and verify specific
> aspects or content.
>
> Obviously this lends itself to neutral scripting techniques with functions
> designed to match the specified functional needs.
>
> CAM is doing this type of approach for business information transaction
> exchanges for machine-to-machine rather than human-to-machine or
> human-to-human content - such as grants submissions example at NIH.

Are there tests within that which could be classed as ODF general
rather than user group specific?

regards

-- 
Dave Pawson
XSLT XSL-FO FAQ.
http://www.dpawson.co.uk

---------------------------------------------------------------------
To unsubscribe, e-mail: 
For additional commands, e-mail: