Summary of 2010-11-09 OpenFormula meeting

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

Summary of 2010-11-09 OpenFormula meeting

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

Attendees:
David A. Wheeler
Dennis Hamilton
Eric Patterson
Andreas G.
Rob Weir
Patrick D.

Regrets:
Eike Rathke


=======

We started here and looked at unresolved issues:
http://tools.oasis-open.org/issues/secure/Dashboard.jspa?selectPageID=10056

At this point, there were some hard ones (unsurprisingly).


* http://tools.oasis-open.org/issues/browse/OFFICE-3435  (HYPERLINK Function)

Rob: There are 4 different suggestions.  E.G.:
- Rob: Just describe computational behavior
- Andreas: Describe computation, note side-effect.
- Dennis: Should be more than a note.  Should be in behavior.
[See comment list for more]

Rob: We could define a pseudo-interface like a 'hyperlink notifier'.
I'd be more confident that we together can work out a proposal.

Dennis: When you evaluate the IRI parameter, it communicates that to the host.

Wheeler: In Excel 2007:
 IF(FALSE(), HYPERLINK("http://www.dwheeler.com";), "http://www.google.com";)
creates a hypertext-linked cell, which links to http://www.google.com
(This is Wheeler's understanding of the semantics provided by Excel 2007.
Basically, if HYPERLINK is *anywhere* in the formula, the cell gets
"marked" as hyperlinking, and the final result of the formula is the URL,
at least in Excel 2007.)
[There were some other sample functions in the chat room, not captured here,
to show how bizarre this function is.]
This is an incredibly weird function.

Rob: Andreas' proposal gets to that.

Dennis: This function just savages everything in the spec.

Rob: Can "may" be used in a note?

Patrick: A note isn't normative, that could be confusing.

Dennis will capture the original proposal as a comment.

Group agreed to Andreas' proposal, but to modify its note.
The new specification is mostly vacuous, and intentionally so.
We're agreeing to make a weak normative statement.  It says what the
function returns, and then just hints that the implementation might
create a URL hyperlink in an implementation-defined way.

Dennis will try to wordsmith the final text of this.



* http://tools.oasis-open.org/issues/browse/OFFICE-3492

Long discussion about what the value for Imperial gallons are (in liters);
we have conflicting authorities.

Wheeler: Just omit "exact", and specify the value already given, and we're fine.

? : What about units not in the official list?

Andreas: The text already says "at least the following units", so it lets people do more.

Dennis: They need a way to add it in a way that won't collide.

?: We could say that any unit name containing "." is reserved for implementation-defined units.

Wheeler: Add statement that "Any unit name containing '.' is implementation-defined."

Eric: That's fine as long as additional units don't HAVE to have ".".

Wheeler: Agree.  Almost all implementations have other units without ".".  The text as proposed does that, I think.

Patrick: What if some unit with "." becomes popular?

Wheeler: Then we could add THAT one, even with a ".".  We're not promising that unit names with "." will ALWAYS be implementation-defined, just that today, any name with a "."
is implementation-defined.  We could add it to the list later if desired.

Resolved.


* http://tools.oasis-open.org/issues/browse/OFFICE-2710

Really a part 1 issue.



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

Eric: Just make the second sentence a note, doesn't need to be normative.

Wheeler: Agree.


* http://tools.oasis-open.org/issues/browse/OFFICE-3493

?: This is a Part 1 issue.

Eric: Yes, but this is the only group who cares :-).

Dennis: Eike's complaint is not right, these are two different circumstances.
I believe they need to be in both places, and need to look at style-condition too.
Need to double-check.

Dennis will double-check this one, and also look at style-condition.



We will plan to meet next week, and cancel if it's not necessary.

--- David A. Wheeler