Warren, your suggestion looked to me like a good way to address the specific problem of row/column formatting that I had described in the proposal.
An optional grid-size element may also be useful in other circumstances, including scenarios we haven't considered. I'd like to revise the proposal to include the grid-size element as well as attributes for formatting an entire row or entire column. Does anybody see any problem with that?
Andreas and Eike, regarding formula references, I'm not sure what the best solution would be to the scenarios you have described. The grid-size element would give consumers enough information to know whether the spreadsheet was created on an implementation with a different grid size, so that they can do whatever they feel is appropriate. Perhaps we could look at some recommendations as Eike suggested, based on use of this element.
Regards,
Doug
Original Message-----
From: Eike Rathke [mailto:erack@sun.com]
Sent: Friday, November 21, 2008 8:56 AM
To: office@lists.oasis-open.org
Subject: Re: [office] Data Grid Size element proposal
Hi,
On Thursday, 2008-11-20 10:34:17 -0700, Andreas J. Guelzow wrote:
> I am wondering whether something like table:grid-size doesn't also
> have repercussions on evaluation of formulae. Imagine a formula that
> refers to a cell location outside the current grid. What happens is
> this formula suddenly finds itself in a larger grid?
Might evaluate to something (maybe something wrong) then, where before it should had resulted in an error. Unless you artificially limit the sheet size as well.
There's more to it though. For named expressions (ranges or formulae) there is the strange and awkward behavior required for interoperability that for relative references the references _wrap_ around the sheet if the position of the formula cell moves towards an edge. When loaded in another grid size, the wrapped position will be different, thus the formula result may as well.
We should at least give some recommendations what to do in such cases, better would be clear definitions.
Eike
--
OpenOffice.org / StarOffice Calc core developer and i18n transpositionizer.
SunSign 0x87F8D412 : 2F58 5236 DB02 F335 8304 7D6C 65C9 F9B5 87F8 D412
OpenOffice.org Engineering at Sun: http://blogs.sun.com/GullFOSS