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

# office-formula message

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

Subject: Re: [office-formula] (from Bryce L Nordgren) Fwd: Use case; multi-radix& roman numeral formatting

• From: "David A. Wheeler" <dwheeler@dwheeler.com>
• To: robert_weir@us.ibm.com
• Date: Mon, 21 Aug 2006 18:06:54 -0400

robert_weir@us.ibm.com wrote:
>
> I like the idea of a formatting in a different number base.

Me too, though I suspect that's a "future extension" rather than
something we
need to do right now.
>  But we would need to define what this means for non-natural numbers.
>  How is -1 displayed in binary?  Two-complement?  Or just -1?
You can't do two's complement unless you have a fixed number of bits,
so "-1" is probably the only safe answer.
>  What about decimals?  Binary-coded decimal or IEEE floating point?

I don't understand this, what do you mean?

> I think Roman numerals leads to other questions.  How do you represent
> numbers greater than MMM ?  How do you represent decimals?  How to you
> indicate zero?  The word "Nullus"?

I'd propose "0" as the representation for zero, it's certainly
unambiguous, and
again allow a leading "-" for negative numbers.
I think applications should be free to display some sort of out-of-bounds
indicator if the number is too large.

For fractions/decimals, I have no idea.

This is in some ways simple; just declare
"this format calls ROMAN() and displays its result".

> We'd also want to keep in mind that the set of allowable output
> formats also sets expectations for parsing of input and automatic
> string to number conversions.

Yes.  That may be the biggest problem, managing expectations.

--- David A. Wheeler

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