Re: [office] Excel 2007 != Ecma spec YEARFRAC. Not even slightly. What should we do?

["Warren Turkal" <turkal@google.com>]

On Thu, Apr 17, 2008 at 2:18 PM, David A. Wheeler <dwheeler@dwheeler.com> wrote:
> > I don't think that we should standardize the excel algorithms What if
>  > we just have a registry of compatibility bases that is maintained
>  > separately from the ODF spec? We could just reserve 5000 basis numbers
>  > that represent a compatibility bases and a 200 or so for a private
>  > area that will never be allocated in the registry for experimentation.
>
>  The idea of having a separate "basis registry" has merit.  At the least, it means we could complete the OpenFormula spec, and point out to a separate registry for that part of it.
>  I think it _would_ be wise to spec the Excel algorithms, because there are people
>  who count on them; but if they're in a separate registry, that makes things easier.

I personally liked your idea of shipping formulas that we are sure
about and shipping others later. We can even reserve bases 6660-6669
for Excel's interpretation of bases.

>  It doesn't address the problem that different people may want
>  basis==0 to mean different things.  The only solution I see for that is to
>  attach some parameter to the sheet (as I noted earlier), which can affect the
>  mapping of the basis value to the algorithm used.

Yuck!....Why can't we just rely on the import filters in apps just map
to the correct bases on import. We certainly shouldn't standardize
magic variables that affect seemingly unrelated formula calculations.
As long as we document it clearly, there's no reason that app
developers cannot handle the conversion.

wt