Re: Open Office file formats (Oasis-open) and gnumeric

[Jody Goldberg <jody gnome org>]

On Wed, Mar 12, 2003 at 08:25:48AM -0500, Russell McOrmond wrote:
> On Tue, 11 Mar 2003, Jody Goldberg wrote:
>
> > implementable.  As an example,  I'm experimenting with a form of
> > conditional formats that is different from MS Excel.
>
>   I think one of the differences is that I think of a spreadsheet as a
> tool used on a network, and you may be thinking of it as a tool used on a
> desktop.  Will users know when they are using features that will be lossy
> when exported to formats other than the Gnumeric format?

Spreadsheets sit in limbo between end user applications and
development tools.  To some people they are just viewing a document,
possibly tweaking a value or two.  In that niche MS Excel dominates
it has had more time and more people beating on it than any of us
can dream of.  However, for several niches spreadsheets are used
very differently.  They are rapid application development and
deployment frameworks.

    - In finance as front-ends to analytics
    - To build schedules
    - As sick simplistic data bases

In these areas features do matter.  If people can get things done
more smoothly with Gnumeric by using calculated formats rather than
hand crafting a maximum of 3 conditional formats when doing a
schedule they will use Gnumeric.

>   Is this feature worth the "best viewed by" problems that will exist?  

Yes.  That is why people are using Gnumeric.  What happens to data
in cells past the maximum size of the application (eg 32k rows in
OO) ?  Sure it can read the data, but it loses it.  How will people
react to recalculating a sheet in a different spreadsheet and
getting a different result ?? Having a shared file format is not
going to solve difference in implementation.  

>   Can you imagine a small team of people trying to innovate by changing 
> the TCP/IP stack outside of a standards body?  It may not be too long from 
> now when innovating on office productivity communications/file formats are 
> looked upon in the same way.

This is the core of our disagreement.  As argued elsewhere I don't
see office file formats as being equivalent to protocols yet. 

>   Excel is a branded tool from a third party, and to suggest it is not a
> standard is overkill (that third party is a convicted monopolist).  Even 
> a renamed snapshot of the current format would be more likely to consider 
> a standard.  If the OASIS Open Office TC ends up with an XML variation of 
> the same functionality, then they will have served their purpose.

An xml based file format with an open sourced reference
implementation is certainly a big step.  It will be much easier to
translate to and from.  However, I don't see any day soon when
people will reliably be able to view a file in a non native app and
see exactly the same thing.  It'll be close, and we've had to sweat
blood to do it for MS Excel, but it does not seem likely to be
identical.