Re: AccrInt

[Jody Goldberg <jgoldberg home com>]

On Mon, Jan 29, 2001 at 11:26:38AM -0600, John Hasler wrote:
> Phillip J Shelton writes:
> > I don't intend to un-Excel clone gnumeric.  If the spreadsheet functions
> > need errors coded in then that is what has to happen.
>
> Then I will neither use nor recommend gnumeric.  I have no objection to
> bug-for-bug compatibility as a _option_, but deliberately coding buggy
> default behavior is a horrendous idea.

Please people read our responses before making these sorts of
inflammatory statements.

The goals of Gnumeric are to provide (in order)
    1) the best possible spreadsheet
    2) a drop in replacement for commercial packages (not just XL)

Notice the order.

Computational differences from XL are fine if they are caused by
improvements in methodology or accuracy of an implementation.
On the other hand bugs in an XL style function are handled by
producing TWO functions.
    - a correct version 'gnumeric_<func>' 
    - a bug compatible version '<func>' for use by the XL importer.

As the import framework is extended we will offer a setting
    'Import XL using bug-for-bug compatible functions'
until then XL import will default to using the buggy version.