Summary 2010-10-12 of OpenFormula meeting

["David A. Wheeler" <dwheeler@dwheeler.com>]

Summary 2010-10-12 of OpenFormula meeting

(As always, please reply-all with corrections.)

Attendees:
David A. Wheeler
Andreas G.
Eric Patterson
Dennis Hamilton
Eike Rathke
Patrick D.

Known Regrets: Rob Weir

Starting here:
http://tools.oasis-open.org/issues/secure/Dashboard.jspa?selectPageId=10070

ISSUES DISCUSSED:

* http://tools.oasis-open.org/issues/browse/OFFICE-2014

Rob Weir was not here, but we don't see anything to discuss.
The technical issues seem to have been resolved.

* http://tools.oasis-open.org/issues/browse/OFFICE-2258 (French accounting).

Assigned to Eric P.  "We have not gotten requests to change this function,
so either Excel is correct, or it's so wrong that no one uses it and no one needs it."

Andreas G.: This definition is different than what Excel does.
E.G., Between 4-5 years, it provides an error.

Eric: Unable to find other authoritative sources.

Wheeler: I don't want to spec something that's wrong; prefer to drop it.

Patrick D: Could we sample public data?

Wheeler: It wouldn't be representative.
Let's wait a week, and if we can't get it resolved, drop this function.
Implementations can always exchange it as an extension.

Eric: I'm okay with that.

Patrick: Does anything depend on this function?
Wheeler: Don't think so.
Eric: AMORLINC [somewhat] does, but that definition is straightforward.
Wheeler: That one isn't under debate, is it?
Patrick: If everyone's happy with AMORLINC, then fine.
Dennis: Does definition in AMORLINC stand without that phrase? Just make that a note?
Wheeler: It's still useful.
Patrick: For AMORLINC, leave in the summary.

RESOLUTION OF SORTS:
Wait 2 weeks to get more info, then if nothing more, drop AMORDEGRC function,
leave AMORLINC there but simplify its description possibly to remove "French accounting system" phrase (leave its summary as-is so people can find the function).


* http://tools.oasis-open.org/issues/browse/OFFICE-2751
"5.8: Need syntax for external references to unsaved documents"

Eric will work to get more info, in progress.


* http://tools.oasis-open.org/issues/browse/OFFICE-2893 (NUMBERVALUE)

The group examined the proposal and found a number of problems.
After some work, we ended up with a revised proposal that all agreed made sense.

Wheeler: It should be DIVIDED by 100 for each %, not reduce (subtraction).

Eike: We may be allowing noise characters instead of failing.

Eike: Need to remove all group separators BEFORE switching to full stop.

The issue here is that this accepts some numbers that might be considered ill-formed:
1,000.555,666 in US conventions (with "." as decimal separator).
Changed rule to fix this.

?: As-specified it's wrong, the first two need to be swapped.

Wheeler: I think "left/right" are ambiguous, but "beginning"/"before" and "end" are not.

Group ended up with a revised and improved algorithm description, which
is now in JIRA as the proposal.


* http://tools.oasis-open.org/issues/browse/OFFICE-3029

Dennis to create a proposal.

Big discussion ensued about proposed the proposed function CURRENT_CONTENT().

Dennis: "It's already there for table:condition".
This is being proposed to be added (so these could be merged).


* http://tools.oasis-open.org/issues/browse/OFFICE-3042

Eric: "I need to work on this a bit more, will do that".

* http://tools.oasis-open.org/issues/browse/OFFICE-3340

(We agreed what to do last time, i.e., it could return Error,
but Eric needs to write up the proposal.)


* http://tools.oasis-open.org/issues/browse/OFFICE-3468

This was a long discussion about IFNA, esp. if it references a blank cell.

Eric: Maybe we can just remove this discussion.

Dennis: It returns either the first or second argument.

We need to check/fix IFNA and IFERROR; consider what happens if a referenced cell is blank.
Similarly with ISNA.

At this point we ran out time.

Wheeler: Let's continue working on this (e.g., on the list).



We will meet next week.

--- David A. Wheeler