openoffice-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rob Weir <robw...@apache.org>
Subject Re: Calc behavior: result of 0 ^ 0
Date Sun, 10 Feb 2013 15:30:05 GMT
On Sun, Feb 10, 2013 at 8:57 AM, Hagar Delest <hagar.delest@laposte.net> wrote:
> Le 10/02/2013 00:43, Rob Weir a écrit :
>
>> Spreadsheets are used by businessmen and not only mathematicians.
>
>
> I think that very few mathematicians use AOO at all.
> Even in the industry (in which I work for almost 15 years), MS Excel is not
> used to perform high level calculations. Even the nice charts we can see
> from complex tests are in fact exports from dedicated software that just
> give the numbers and let MS Excel draw the chart with all the bells and
> whistles needed (secondary grid, axis, legends, ...).
> Real calculation is made with specific applications, often developed by the
> companies to be sure they master the process.
>
>
>
>>> - In 3.4.1, "=0 ^ 0" returns 1
>>> - In 4.0, as patched by Pedro (see issue), "=0 ^ 0" would return an error
>>> - According to ODF, valid results are 0, 1, error
>>
>>
>> In other words, the results we were giving before were entirely valid.
>
>
> It just means that an acceptable shortcut was used. Giving 'error' would be
> valid too (regarding the ODF compliance).
> But I would not say that mathematics POV 1 as a result is valid. Even if
> some tend to think that 1 is OK.
>
>
>
>> Microsoft has gone decades with treating the year 1900 as a leap year.
>>    Should we?
>
>
> Agreed, who cares what MS Excel does? It should not dictate what AOO should
> do.
>
>
>
>>> - We lose backwards compatibility if someone was relying on the fact that
>>> OpenOffice returns 1 as the result of "=0 ^ 0"
>>
>>
>> Correct.  The fact is we have returned 1 for this calculation for over
>> a decade.  Whether mathematicians think it is right or wrong (and they
>> do not all agree), that is what we did.  So changing it now has the
>> potential to break real user spreadsheets. So this is a serious
>> change.
>
>
> First, how many users would face such a problem when both the base and
> exponent are null???
> Again, real maths are not done in spreadsheets. So it's very likely the 0^0
> cases would not break that many sheets and with an error, users will be able
> to spot quickly the issue and adapt to prevent that situation to be
> calculated.
>

It does not follow that if "real mathematicians" do not use
spreadsheets, that making breaking changes does not have consequences.
 That logic is nonsensical.

And again, it is mistaken to assume that the person who suffers
because of a newly introduced error is the person who can correct the
error.  Just as with extensions, the person who designs a sheet and
the person who uses the sheet might be different people.

> Second, and that's my main point here: you're angry about such a minor
> change when you don't mind breaking the backward compatibility of the whole
> extensions eco-system? See:
> http://www.mail-archive.com/api@openoffice.apache.org/msg00107.html
> I'm lost about the priorities... and how end-users fit in your agenda.
>

Don't put words in my mouth.  I'm not angry.  And when did I ever say
I was happy about breaking extension compatibility?  Quite the
opposite.  I've called for all breaking changes to be proposed and
reviewed on the dev list.

-Rob

> Hagar

Mime
View raw message