Re: why>?

I disagree.

if you spend an hour a week inside of excel you should be classified as
a 'software developer' and you should be held to the same standards of
excellence as real developers.

you should have to use visual source safe.

you should have to have 9 different managers

and your *** should be outsourced to india.

you guys are the ones that are easily replaceable.

I replace your work all day long; it is my mission im life to put
spread*** dorks out of work.

Again Harlan; i've never had any problems with floating type in
databases.
i haven't ever had a problem; in any way shape or form.

My problem comes when im import a field that should be MONEY and some
FUCKTARD puts the dollar figure 'DONT REMEMBER'

or 'I DONT KNOW'

or 'CUSTOMER DOESNT KNOW'

in a database; i can protect against dipshits like this.

with excel it is impossible to ensure datatypes.



-Aaron


Harlan Grove wrote:
aaron.kempf@xxxxxxxxx wrote...
I've never had a problem with float data types in anything other than
excel.

Probably because you've never checked your floating point carefully.
Try this. DBMS table T with a floating point numeric field F containing

F
0.375
0.375
1.375

All are exact binary and decimal values. Now set the numeric format to
fixed with zero decimal places. This field will appear as

F
0
0
1

and the sum with the same number format will be 2.

Fixed fractional precision is a problem that affects equally ALL
software capable of performing arithmetic with fractional values, SQL
Server and Excel alike.

excel just has a crappy implentation of it.

No, far more likely you don't know how to anticipate when it'd be a
problem, and you're ignorant of the standard techniques for handling
it.

you're not a software developer?
...

Not acording to my job description, no. While I've built and maintain a
few different models (a few in Excel, a few others not much more that
batch files and VBS scripts, and a few compiled .EXEs), it's not the
central focus of my job.

anyone that keeps excel open more than an hour a week is a software
developer; but you slapdicks aren't held to the same standards of
excellence.

More incoherence. Anything human-readable generated by a computer is a
'report', and anyone who has Excel running more than 1 hour a week is a
'software developer', even if said person just performs data entry.

I'll grant that too many businesses use too many spread*** models
built by people with no experience in software engineering and no clue
how to manage software development projects. But that's not the same as
saying all spread*** use is like that.

However, many people still use computers to compute, and very few of
them would be considered software developers under any reasonable
definition or job description. That I drive a car doesn't make me a
professional driver. That I cook meals at home doesn't make me a chef.
That I give people directions when they ask me on the street doesn't
make me a travel agent. That I write formulas in spreadsheets doesn't
make me a software developer.

if there were a certain level of basic skills that you spread***
dorks have-- then maybe it wouldn't be so difficult.

Again I return to the calculator metaphor. Spreadsheets when used as
glorified calculators require no more certification than more mundane
calculators. They're just tools for performing repetitive calculations.

but as it is each spread*** developer has this misguided concept that
their applications are 'so unique'

In terms of the calculations, one time series forecasting model SHOULD
BE similar to any other, so should any discounted cashflow model, so
should any mocked up pro forma financials, etc., and usually there's
only one such model where a sinlge authority is in charge. Different
departments may have similar models in use, and the similarities may
outnumber the differences, but it'd be a political nightmare to
consolidate them. The reasons there are so many variations on the
common themes has little to do with the deficiencies of spreadsheets
and much to do with interdepartmental turf wars.

if you had to take classes in excel; in order to use excel
maybe excel wouldn't be such a waste.
...

You mean the classes you so obviously missed given the apparent gaps in
your Excel knowledge that have already come to light? Yes, I agree that
someone with a mind as tightly closed as yours should be required to
demonstrate CLAIMED proficiency.

some sort of standardizing training really might make excel a usable
technology.

It's not training in Excel per se that's needed, it's training in
general software engineering techniques - revision control, regression
testing, specification, and documentation.

how many million-dollar spreadsheets have you worked on buddy?
...

Mercifully none. Nor would I try. Only fools and Microsoft marketing
shills would claim Excel should be used for such things.

That said, I should point out that there were mistakes in financial
statements in the paper, pencil and calculator era, but back then
managers took some responsibility for checking their subordinates work
for reasonableness. Spread*** horror stories are more depressing for
what they imply about the state of corporate management than the
deficiencies of spreadsheets as modeling tools.

One case in point, at least in terms of my own responses in Excel
newsgroups, is the general overuse of ISERROR by others. ISERROR should
never be used in Excel. It's an invitation to BIG TROUBLE. There are
situations in which one needs to filter out #VALUE! and #N/A errors,
because Microsoft too slavishly followed Lotus Development Corp's specs
for such functions as MATCH and FIND for which there was NEVER ANY GOOD
REASON to return error values rather than 0 or -1 to indicate 'not
found'. But there's **NEVER** a good reason to filter out #REF!, #NAME?
or #NULL! errors. #DIV/0! and #NUM! are problematic. But it's possible
to use ERROR.TYPE to filter out only specific errors. That's the advice
I give.

i was automating pulling information out of excel while you were still
in diapers so don't you talk to me like that; in that condescending attitude.
...

Your brain is in the diaper pail, or your way off figuring my age. But,
then, you've already demonstrated that you have a hard time with
numbers, so maybe you can't figure out which is older, Excel or me.

I'll drop the condescension when & if you DEMONSTRATE any Excel
competence. It's easy for you to claim such a wealth of knowledge &
experience, so maybe you're just intellectually constipated when it
comes to letting any out.

your stupid silly spreadsheets can't even be used by 2 users at the same time.

Incorrect. They can be used by an arbitrary number of users at the same
time, but only one user may save changes to any one copy at the same
time. Spreadsheets don't make sense for data acquisition systems even
when they send their entries to databases. It's make more sense to use
database forms in the first place. On the other hand, spreadsheets make
a lot of sense when there are relatively few user entries but hundreds
or thousands of calculations based on those entries. For ad hoc
calculations, in which there's no real need to store anything other
than the model results, and then only as virtual or actual printouts,
none of the users would need to save anything. In that case, millions
of users could load and run the same spread*** because none would be
saving any copies. That's what I mean by properly conceived spread***
models.

your silly spreadsheets make you copy and paste and copy and paste--
it's ridiculous that there isn't more to your life than copy and paste.
...

Have fun with Google Spread***. It'll make you do LOTS MORE copying
and pasting than Excel.

.

Loading