To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=92153
------- Additional comments from [EMAIL PROTECTED] Mon Aug 11 23:19:22 +0000
2008 -------
Quote discoleo:
The problem with the proposed approach is:
1. variable names have to be typed
***What do you mean? How will Calc know I want to name a variable 'Abcd02Jimmy'
unless I type it in?
2. cells can be selected with the mouse
I mention this, because it is easy to mistype something like AB_6 or AB6v and
still end with cell [AB6] instead of the variable name.
***There seems to be a misunderstanding. Variable AB_6 has nothing to
do with
cell AB6. it could be associated with cell H41. (Why I call a variable 'AB6' is
my business; it could be because it contains the product AB*6, or whatever...)
In fact, that's the problem: if I use 'AB6' in an expression, Calc assumes I
refer to cell AB6 when I really wanted to refer to cell H41.
IF the user names something PI2, or to give an example that I often use APACHE2,
then I still believe that the chance of an error will be dim and the chance of a
calculation with a cell PI2 (and APACHE2 - when spreadsheets will support as
many columns) is rather small.
But as you said, it is not zero.
***Exactly! To me, 'PI2' referred to Pi*2 (ie, 6.28...) which I had
pre-calculated and stored in cell D1; Calc (ver.3) assumed I was referring to
cell PI2, which is empty. But if the cell were'nt empty, Calc would not give me
a div0 error, but it would quietly give a wrong result!
Instead of forbidding valid names (and highly
desirable names), I would adopt an overall better solution:
- Calc could detect such conflicts and mark the respective cell
-- mark only IF the cell is non-empty
-- it is unlikely that the user would reference an empty cell
***'unlikely', 'dim' or 'small' chances (above) and 'rare occurrences'
(below)
are not a good basis for reliable programming. What if the non-empty cell
becomes empty later? Or, if the user references an empty cell he intends to
fill?
I still believe that the liberal use of names greatly outweighs the negative
consequences. If people were using more often names, then they would start both
to organise better their spreadsheet and also reduce a great number of other
errors. This would by far offset the rare occurrence of the mentioned naming
conflict.
***'liberal'? There are always restrictions to variable names: reserved
words
like FOR, IF, ELSE in many languages, forbidden characters like space, ?, ., &,
etc. So, forbidding the family of reserved words of the form AAAnnn (eg, AB6 or
APACHE234) should surprise or upset no-one, and would prevent confusion with
cell references once for all --no 'ifs', 'that's unlikely', 'that will seldom
happen', ...
Incidentally, you mention the problem of typing mistakes. Well, it's
easier to
type 'MS' instead of 'MonthlySalary' or 'monthly_salary', but the later forms
are recognized as far superior for clarity. The meaning of 'MS' may be obvious
when you write it, but will be far more obscure when you revise the spreadsheet
6 months from now. Unless you reserve a table of comments where you explain that
'MS=monthly salary'-- but that has to be typed!
---------------------------------------------------------------------
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
