David A. Wheeler wrote:
> > Actually, how about IFERROR and a new function, IFNA?
Eike Rathke:
> I like the idea behind IFNA additionally to IFERROR to distinguish this
> case.
Excellent! Unless there's a complaint ASAP, I will add drafts for these two functions (IFERROR and IFNA). Since MOOX has the default value as "", I intend to do the same for IFERROR, and for consistency do that with IFNA as well.
> In the current state of the spec only LOOKUP, VLOOKUP and HLOOKUP
> are defined to return NA() though if the lookup value isn't found.
>[and FIND and SEARCH...]
> What other functions would we need to define to return NA() on some
> condition?
Well, there's always NA() :-). Seriously, I think that's enough. We have a number of search functions that return NA(), and now we can easily change what NA does using IFNA(). Similarly with general errors.
> I think IFEQ would be only confusing..
Okay. I think the utility of IFEQ goes _way_ down once IFNA() and IFERROR() are included, so let's drop IFEQ().
--- David A. Wheeler
