[rfc-i] [IAB Trac] #266: Requirement for "Clear Printing"

Joel M. Halpern jmh at joelhalpern.com
Tue Feb 19 06:30:50 PST 2013


As far as I can tell, there are two fine-grained reference issues.  One 
is for machines (anchors).  I am less concerned about that, and if it is 
an issue it is unrelated to clear printing.

The other is for humans.  It seems to me that fine grained references 
need to be independent of whether you are looking at a printout, an 
HTML, a PDF, etc.  And given that we are talking about allowing multiple 
canonical forms and expecting folks looking at different forms to 
converse, it seems to me that whatever we expect to use for fine grained 
human references has got to work across formats.  This seems to rule 
page numbers completely out of the picture for this problem.

Thus, while we do need the ability to easily print clear, readable, and 
usable copies of our documents, that does not seem to be related to our 
need for clear and effective fine grain references.

Yours,
Joel

On 2/19/2013 3:16 AM, Brian E Carpenter wrote:
> Joe,
>
> On 18/02/2013 17:13, Joe Hildebrand (jhildebr) wrote:
>> On 2/17/13 1:17 AM, "Brian E Carpenter" <brian.e.carpenter at gmail.com>
>> wrote:
>>
>>> On 16/02/2013 22:49, Joel M. Halpern wrote:
>>>> Conversely, why should the canonical format be required to support page
>>>> numbers?
>>> I think Ran already answered that - for unambiguous ease of reference
>>> (especially if a document happens to contain very long sections that
>>> span more than one page).
>>
>> If you made the requirement "capability for fine-grained references", you
>> would be speaking requirements language instead of implementation language.
>
> OK, what I mean (and I suspect Ran too) is:
>
> "capability for fine-grained references when working with a printable format."
>
> That's old-fashioned, I know. However, you are advocating fine-grained
> references via anchors, which is:
>
> "capability for fine-grained references when working with a hypertext format."
>
> The question before the house is whether we want both of these, or only one.
>
> Hmm. This is the same discussion I must have had in 1980 with TBL, when he
> showed me the Enquire program as an alternative to software documentation
> printed on a Diabolo.
>
>      Brian
>


More information about the rfc-interest mailing list