OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

office-comment message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [office-comment] [Formula SC] Treating errors as strings?


Hi Brad,

Some overdue answers to your comments.

On Thursday, 2006-12-21 10:57:16 +1100, Brad Hards wrote:

> I am confused by what section 4.5 of the OpenFormula specification is trying 
> to say.
> 
> If the result of a calculation results in an error, then the result of 
> calculations on that result either could propagate the error, or treat the 
> error as a string. 

Please note that the wording about string refers how the error result is
stored in the OpenDocument (ODF) file format. This is not to be confused
with the formula content or calculation or result. A calculated formula
propagates the error to dependents.

> For example LEN(1/0) could be Error (or #VALUE!, however the error is 
> presented), or it could be a integer (5 or 7, depending on the length of the 
> string). 

No, it can't be an integer result, the result will be an error value.

> Perhaps Section 4.5 should say that "calculations taking inputs that are Error 
> values shall result in Error, except where a specific exception to this rule 
> is defined in the formula or operator part of this specification".

Nice wording, I'll weave that in.

> This would avoid nasty cases where calculations hide underlying errors, as 
> implied by Section 10.1.

That is not about error result values. It is about miscalculations that
may happen and rounding errors and errors people make when setting up
spreadsheets for their calculations, assuming false prepositions. The
section IMHO should also not be part of the final specification, maybe
in an annotated version.

> Also, Section 4.5 of version 20061208 has a spello (propogate should be 
> propagate) and an incomplete sentence: In OpenDocument, if an error value is 
> the result of a cell computation any must be stored that way, it is treated 
                                     ^
> as a string. 

Should read "and must be stored" instead. Thanks.

  Eike

-- 
 OpenOffice.org Engineering at Sun: http://blogs.sun.com/GullFOSS


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]