Re: [office] Formula Sub Committee Draft Charter

[robert_weir@us.ibm.com]

That's fine with me.  Whatever
we end up with with needs to be formally specified, so unless we define
the syntax of BNF in a self-contained way, we'll need to refer to an external
standard.  Whether it is XML 1.1, or RFC 2234 isn't as important to
me as the fact that we're following some authority.  

-Rob

"David A. Wheeler" <dwheeler@dwheeler.com>
wrote on 02/03/2006 01:05:40 PM:

> One quick comment on the grammar stuff:
> 
> > > Suggest the grammar is done in ABNF form according to IETF
RFC 2234.
> 
> I suggest NOT adding this to the charter, but instead let the group
> discuss that & come back with a recommendation to the TC.
> 
> In my opinion, IETF RFC 2234 would not be the best choice for the
> BNF.  2234 is optimized for handling email, which isn't the 
> application in view,
> and its notation is NOT what most people use for BNFs.
> For OpenFormula we used the BNF format defined in XML version 1.1.
> Since formulas are normally inside XML documents, this made
> more sense... it's what most people who use XML are familiar with.
> I think it's easier to read as well.  And this BNF is ALSO
> a defined standard (by the W3C in this case).
> 
> Maybe there's a better choice, or maybe 2234 is better for
> reasons I'm unaware of. But that's the sort of thing best left to
the
> subcommittee to hash out, and recommend back to the TC,
> instead of cutting off discussion ahead of time.
> 
> --- David A. Wheeler