[rfc-i] draft-flanagan-rfc-framework-02 composition tools

On Mon, Nov 03, 2014 at 07:39:18PM -0500, Joel M. Halpern wrote:
> Having used MS Word on some large I-Ds I have to say that I am getting tired
> of the assertion that the only folks for whom Word won't work are the one
> Joe cites below.

Indeed, I find Word to be the wrong tool for every job.

But anyway, Word is _certainly_ not even possibly the long-term
archival format, because "Word" isn't a format but a localized
reference to whatever format happens to be generated by the set of
tools one has to hand.  I'm aware of the Office XML format, but
there's still as I understand it no clear and obvious way to restrict
any given version of Word to any specific subset of those XML features
(or if there is, my Word-fu is too weak).  Moreover, because of
various IPR issues, non-MSFT tools don't interoperate perfectly,
making an even bigger mess. 

So what we're really talking about here is editor preferences, and I'm
not exactly sure why we are permitting the two things to be conflated.
We long ago decided that the XML format was going to be the "source
document".  I understood then and understand now the arguments against
that.  It seems to me that if one is serious about archival stuff then
the only thing that was ever really an option was PDF/A, and that
wasn't good because of the problems of reflowing, different
page/viewer sizes, and so on.  That led us to the decision for a
source format.  Reverting that discussion to generation-tool choice is
a step backwards.  It seems to me that we shouldn't have that
discussion.

Best regards,

A

  In one case, we found using Word so unweildy that we
> hand-converted it into XML.
> 
> I am not asking the Joe Touch stop using Word.  But the assertions that it
> is workable for the bulk of us make assumptions that are not supported by
> the evidence.
> 
> If the request is that we plan for conversion tools to permit authoring in
> MS Word, with the understanding that it will be less effective since there
> is no way to capture the original metadata cleanly in MS Word, but instead
> the tools must guess the values, I could see us including that in the
> migration plans.
> 
> Yours,
> Joel
> 
> On 11/3/14, 7:29 PM, Joe Touch wrote:
> >
> >
> >On 11/3/2014 2:54 PM, Ted Lemon wrote:
> >>On Nov 3, 2014, at 4:50 PM, Joe Touch <touch at isi.edu> wrote:
> >>>I agree, but why do we want to use xml2rfc if it boxes us into things
> >>>like this?
> >>
> >>You saying this as if you were proposing some alternative that would
> >>box us in less.   But you haven't made a real proposition--all we
> >>have is that you think Word would work.
> >
> >I want a solution for which Word can be used as an authoring tool.
> >Support for authoring tools hasn't been a key factor in the progress to
> >date.
> >
> >>But we know that Word
> >>doesn't address a great many of the very definite requirements we
> >>have articulated; in particular, it would not be usable for some
> >>significant percentage of the IETF membership.
> >
> >The percentage of the IETF that does not use Windows, MacOS, or Linux,
> >agreed.
> >
> >>And it boxes us in far worse than xml2rfc does.    You said we should
> >>aim higher, but we *are* aiming higher than Word.   It just depends
> >>on what you mean by "higher."   You are putting usability for
> >>Microsoft Word users as the highest value, and trying to drag
> >>everything else through that wormhole, but first, that's not our user
> >>base, and second, usability for document editors is not the only or
> >>even the most important requirement for this solution.
> >
> >I have seen a lot of requirements driven from the need for automatic
> >extraction of document structure or contents, but have not seen a good
> >rationale as to the need to do that from documents.
> >
> >>So this is why you aren't getting the traction that you think you
> >>should be getting here: you are pushing a solution to the wrong
> >>problem.
> >
> >The problem is DOCUMENTATION.
> >
> >This isn't a programming project.
> >
> >Joe
> >_______________________________________________
> >rfc-interest mailing list
> >rfc-interest at rfc-editor.org
> >https://www.rfc-editor.org/mailman/listinfo/rfc-interest
> >
> _______________________________________________
> rfc-interest mailing list
> rfc-interest at rfc-editor.org
> https://www.rfc-editor.org/mailman/listinfo/rfc-interest

-- 
Andrew Sullivan
ajs at anvilwalrusden.com