[rfc-i] re: new MS Word template

Joe Touch touch at ISI.EDU
Sat Oct 30 15:50:16 PDT 2004



John C Klensin wrote:
> 
> --On Fri, 29 Oct 2004 17:23:24 -0700 Joe Touch <touch at ISI.EDU>
> wrote:
> 
> 
>>Hi, all,
>>
>>I have just completed an update to the Word template from RFC
>>3285. It  includes a number of changes, including:
>>
>>    o  redefines basic styles (Normal, Heading1, etc.) rather
>>than        creating new styles (enables auto-renumbering in
>>outline mode        when demoting/promoting)
>>
>>    o  updates boilerplate according to RFC 3668
>>
>>    o  uses more conventional methods for autonumbered
>>references and        figures
>>
>>The template and post-processing code, and Internet Draft
>>describing  this new version are available at:
>>http://www.isi.edu/touch/tools
>>
>>(the ID will be published after the IETF).
>>
>>Comments and input appreciated.
> 
> 
> Joe,
> 
> Based on discussions the last time around, this may be the wrong
> thing to do.  Or perhaps not.  Things for you to consider...
> 
> (1) While redefining the basic styles will undoubtedly make
> things work more smoothly, those types of changes have a
> tendency to be irreversible.  That is, for those of us who might
> want to use Word for RFCs but who must use it for many other
> types of documents according to organizational norms, your
> redefining basic types is nearly a showstopper.  That certainly
> argues for "fork" in preference to "revision".

The styles are redefined in the template -  use of the template does not 
override these styles for documents based on other templates, e.g., 
Normal.dot, or IEEE .dot files. I've already verified this on my 
machine, using the same machine for RFCs, IEEE docs, and a number of 
others, and have used this sort of 'redefined basic styles' for many years.

More direct evidence that this is a 'showstopper' would be constructive.

> (2) I think that, with the advent of Word 2003 and its near-XML
> output capability the right way to do things going forward
> probably involves building appropriate style sheets and a
> Word-MS-XML to RFC 2629bis translator.   That would give us an
> archival/ editing form we could deal with and that was not
> dependent on particular versions of Word.

That depends on XML support, which depends on a particular version of 
Word (2003). As with nroff or xml2rfc, the assumption is that future 
versions of word would reasonably support this template for a while. 
Rather than start with something that supports 95 and beyond, 98 and 
beyond, etc., I chose 2002, but there's no reason that (a few) newer 
versions of Word should have problems with this template. Again, that is 
my experience with templates.

> It would also deal
> with the line length issues and the almost-ASCII character
> problems with which the Word template approach has never been
> optimal.

Specific corrections would be useful. The current version, AFAICT, obeys 
the line length fine. The issue with non-RFC ASCII characters (not all 
ASCII are RFC-compliant) and extended ASCII characters is harder to fix 
without a post-processor or custom output device. The latter is closer 
to how nroff and xml2rfc fixes things, but is a bit overkill given 
reasonable perl script capabilties, IMO. At some point this 
post-processer scripting can be built into a macro that is triggered on 
print or save.

> If needed, I'd assume that some of us could cobble up
> a Word 98/2000/XP -> Word 2003-> MS XML translator on a server
> somewhere with little trouble.
> 
>      john

That still harkens back to the nroff-style of 'edit, translate, view'. 
My goal is to move more toward something that can print directly to a 
printer, without going through intermediate stages, as well as which 
renders on the screen in something close to how it ends up printign.

Joe
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 254 bytes
Desc: OpenPGP digital signature
Url : http://www.rfc-editor.org/pipermail/rfc-interest/attachments/20041030/a86d8168/signature.bin


More information about the rfc-interest mailing list