Re: why>?

storage isn't a concern anymore harlan

catalogs, contracts, manuals--- all that stuff is easily written using
a real reporting tool.
Personally I think that everyone from Eddie Bauer to lawyers offices--
should throw out everything that they do-- and rewrite stuff in Access.

I mean-- what better way to build a database-driven catalog; than to
use a db-reporting tool? I argue that Access is better and making
pretty flyers than publisher or word.

I argue that I could write a book in Access; and it would be easy to
manage.
Instead of having hundreds and thousands of manual formatting in Word
or Excel--you build a template; and Access generates stuff for you on
the fly.

and the clincher is this.. when you write your contracts usign a db
reporting tool.-- if you need to digest this data at some other time--
instead of typing it in by hand; you've already got it in db format.

say you need to digest these contracts in other mathematical formulas.
Instead of looking up all the details from printed paper; and typing it
into Excel.. you source this data from where it should be stored-- in a
database-- and you leverage it all from one place.



db aren't complex; there isn't a single task in the world that is
easier to do in a spread*** than it is to do in Excel.


i have provided lots and lots of solutions.

you still have this idea that when you cartesian data that the data is
actually cartesianed and then filtered out.

I still think that is the root of our miscommunicaton.

If i cartesian all letters to form all letter combinations-- and then i
filter it to only use the letters A and B-- we can demonstrate that
this doesn't loop through every possible combination and then filter
them down.

it's obvious from looking at the results-- that databases are lightning
fast.

I just think that you should be doing more to automate.
Instead of keeping data in documents; instead of copying and pasting
stuff from webpages-- why don't you source data into a database and
then re-use it??

if you can do it today and you have to re-do it tomorrow-- why not make
a multi-user application where people can all enter their budgeting
information at the same time?

with excel; you can't scale the workload to multiple people.
that is probably my biggest complaint about Excel-- is that multiple
people can't work in the same XLS at the same time.

and I dont have a beef with PDF any longer. Adobe is guaranteed a
future because they bought macromedia.

I personally think that Adobe Acrobat Reader is the worst-written piece
of software EVER. imagine excel-- worthless little program-- that is a
TSR and about twice as bloated.

I mean-- it takes 60 friggin seconds to open?
and then it sits around in my memory?

it is ridiculous.

i mean-- how many times do i have to bring up task manager to kill out
of adobe acrobat reader?

I can produce portable reports in PDF form quite easily.. just as
easily as you. Technically; I can use one Adobe Acrobat license for 20
users and you have to purchase one license per desktop.

To be even more honest-- Adobe Acrobat costs the same as Microsoft
Access; so if push comes to shove; i sure wouldn't spend $100 per seat
to further a monopoly.
I would look for 3rd party alternatives..

a) export reports to SNP format.
b) have a centralized machine 'print' SNP to PDF
c) rinse and repeat.

WOW. Harlan... I just saw this:

Spreadsheets are poor platforms for writing complex
systems that would be run unattended

this right here-- it gives me a 40% more respect for you (meaning your
total respect percentage is about 37% altogether). You actually admit
when Excel has a weakness.

I don't agree with your notion of 'properly concieved spread***
models shouldn't allow email portability'.

I don't believe that you have seperated what is desirable from what is
practical. I think that you are thinking inside the box.

I believe that these 2 concepts go hand in hand-- the ability to
schedule something to run for unattended execution--

and the ability to send emails programmatically.

I believe that the ability for software to run in unattended manner is
a critical -- i mean-- absolutely critical-- requirement.

Have you seen the movie 'Brazil' is it (it might be 1984; I can't
remember)? Rows and rows of people sitting around; pushing
calculators?? Rows and rows of people; even robots-- sitting around and
typing stuff into calculators?

it is hilarious I think.

Right now; you proved that you are tethered to your desktop; tethered
to copy + paste in ways that just aren't problems in the db world.

That is why I say things like 'i get more done by 2am than you do all
day'

Now that you agree with me-- I will drop those portions of this
discussion.

I respect and thank you for continuing to talk these things over with
me.
i just honestly don't understand where people get in the habit of using
Excel.

I just don't understand it-- in the big picture of things.

Is it taught in colleges?


I still believe that everything out there is a report. And I dont
believe that Excel has any functionality for reporting. I don't
believe that Excel has 1/4 of the functionality that I need-- in order
to print pretty reports off of db-driven data.

And I dont think that Excel should be as popular as it is without this
functionality.

If and when Excel has a capability to send a portable report-- that is
when i begin to see a drop of value in Excel. If Excel could spit out
PDFs; that had embedded data-- then Excel would be a dream come true.

(I'm not talking about this crap MDI format-- i mean; microsoft just
ISNT ALLOWED to change portable document formats 3 times in 5 years).

As it is; i just don't understand the value proposition.

It's like-- I build aircraft carriers-- where my mothership databases
can launch little tiny reports-- portable reports that are easy to
format.

And spreadsheets are battleships. big and slow and dumb and
inflexible.

If you know anything about history; you know which type of ship is more
powerful (we lost what; 4 out of our 6 battleships at Pearl Harbor?)

I honestly wish that I understood a lot more about how to minimize file
sizes in Excel.
If i could find a book that is specific to that topic-- then I might
see more value in Excel.

As it is; i just see thousands and thousands of formulas-- that are 80%
the same with different arguments.

And I dont think that things are that complex in the db world.

I don't -- in the db world-- need to write formulas-- for row 273; i
dont need to hard-code '273' in any of my formulas.

I see that you have to do things like that all day long.

So i just see the difference between Excel and databases-- that
databases don't NEED to store a different formula for each row.

And that is why I believe that they are 10,000 times more efficient
than Excel.

If i have a couple of tables (5 tables, 1000 rows each); and I render
these in 30 different slices--

from my understanding..

5 tables
1000 rows in each
100 characters per row -or- 10 columns of 10 characters each.

in a db world-- i have one copy of the data and 30 queries. Total
storage requirements = 5x1000x100 characters per row + 30 sql
statements (100 chars each) =
500,000 DATA
3,000 FORMULAS
503,000 TOTAL CHARACTERS.

in excel -- i have one copy of the data and 30 queries. Total storage
requirements = 5x1000x100 characters per row + 30 worksheets of
formulas (50 chars each formula, ) = 500,000 DATA
15,000,000 Formulas (30 sheets X 10,000 cells X 50 chars/ formula) =

15,500,000.

I just don't understand-- how and why-- your report takes 15,500,000
characters.
and my report takes 503,000 characters.

I mean-- if we include indexes-- I dont index as much as other db
people because I'm an Olap nut (and Olap is effectively the same as
indexing)

So if we do include indexes; and we index the snot out of my db-driven
solution.. then maybe we're stuck with 1,003,000 characters vs.
15,500,000 characters.

I just don't understand-- it boggles my mind-- how verbose Excel is.

Please-- correct me where I am wrong-- I want to get to the bottom of
this..

-Aaron


PS - I do believe that Access reports are a better tool for printing
than Excel. I just feel like the whole 'shrink to fit' feature results
in fonts that are unreadable in many situations-- and instead of
shrinking everything; i wish that Excel had the capability to stretch
data NORTH-SOUTH or even to just shrink a single cell.






Harlan Grove wrote:
dbahooker@xxxxxxxxxxx wrote...
A) properly conceived spread*** models
IS AN OXYMORON.

what do you mean by PROPER?

Models that perform a lot of calculations with possibly a lot of
*static* data and a little user-entered data. One example, something I
built, takes up to 14 entries from the user and uses some of those
entries as weights to produce cashflow patterns which are provided to
customers in PDF files (company policy; if you don't like it, bang your
head against a wall until the feeling passes). Behind the scenes the
calculations involve lookup, interpolation and weighted averaging.
Simple but tedious.

All users do is make the entries and write PDF files. There are 2 free
form text entries (unvalidated - my users are on their own for spelling
customer etc names correctly), 6 dropdown lists (validated to the
extent that if users enter anything not in the lists, all formulas
return errors and messages appear next to those cells explaining what's
wrong), and the remaining 6 are two separate sets of 3 percentage
entries with each set also having a formula equal to one less the sum
of the 3 entries. If any, including the formulas, are nonnumeric or
outside the range 0 to 100%, messages appear beside the formula cells
and all other formulas evaluate to errors.

It's a glorified calculator. Users don't save hundreds of copies of the
workbook, though there is a separate copy on each regional office's
file server. It does generate lots of PDF files, but since these files
are intended for different customers, that's not a problem.

Could this be provided by a web page? Probably, but there'd need to be
some user authentication, and that'd need to be something that would be
able to prevent people who've recently quit working for customers and
gone to work for competitors (a situation that occurs not infrequently)
from accessing this information. Yes, they could take existing PDF
files with them, but they could also take photocopies of other
sensitive material. However, the goal would be to prevent them from
creating NEW exhibits using different entries. So unless our customers'
personnel databases were linked to our user authentication process,
it's safer to keep this offline.

This is something that *IS* easier to do with spreadsheets than
databases. It's POSSIBLE to do it with either, but it's EASIER with
spreadsheets. Try dabbling with interpolation to see why. Given tables
of quarterly cashflow percentages, that is, each quarter has a
percentage between 0 and 100%, and the sum of quarterly cashflows over
all quarters is always 100%, users can select semiannual, quarterly,
monthly and weekly cashflows (with standard simplifications: each
quarter has 3 equal duration months and 13 equal duration weeks).
Semiannual cashflows are sums of two sequential quarterly cashflows,
but monthly and weekly cashflows are linearly interpolated between
quarters. Granted you could store each type in database tables, but
there's a trade-off between storage and calculation time. Besides, the
same thing goes on in databases when views are used rather than tables.

B). I can do excel macros; i've written hundreds and thousands of pages
of them. I dont see the point in trying to knock me Harlan. I know
excel. And I hate it. Is it that hard to understand?
...

I understand the claims you're making, but I don't believe them. Prove
it. I don't doubt you could tweak recorded macros and apply generic
techniques that would work with other applications using VB*. But I'm
unconvinced you know how to use Excel's object model efficiently or
intelligently.

I DONT DO LOW-LEVEL STUFF I DEVELOP INTERFACES FOR DATAMARTS,
DATAWAREHOUSES. I automate 20 servers at a time.

More unsubstantiated claims.

I DISAGREE WITH YOU. . . .

No?! Really?!

. . . Get off your high horse; everything that you do
in excel is repetitive. Everything that you print is a report.

How predictable. Everything Aaron can conceive of is a report. Or
should that be the only thing Aaron understands is reports. So all
computer output is reporting, ergo all computer usage is report
generation.

. . . Contracts are nothing more than reports; retard. . . .
...

Predicted that response, didn't I?

Only you would asert such an absurdity. Only you might believe it. Gee,
I hope you have the opportunity to see whether this is true or not.

Next you'll be telling us all marketing material and advertising are
reports. After that, all literature is reports. Anything that could be
[re]produced with a computer and a printer is a report. Proof via
reductio ad absurdum that you're an idiot.

I DISAGREE WITH YOU. Databases aren't unnecessarily 'complex'.
Databases don't need IT. Support Infrastructure-- Warning
Mechanisms... Just because you had an email fail once; is it the
databases fault?
...

More manufactured BS. When did I claim any e-mail failure? I don't see
the point to e-mail notification for PROPERLY CONCEIVED spread***
models. To make this clear, I believe spread*** models should be
interactive. Spreadsheets are poor platforms for writing complex
systems that would be run unattended. Even for models in which the
unattended runtime would just be very long recalc, if spread***
models take more than 30 seconds to recalc, they shouldn't be
implemented as spreadsheets.

If you're right that databases aren't unnecessarily complex, wouldn't
your time be better spent showing poor benighted Excel users how to
solve their problems using databases? That is, if your goal were
convincing Excel users to switch to databases. There have been annoying
Perl solutions in many other languages' newsgroups over the years, so
off-topic solutions have a USENET tradition even if they are contrary
to proper netiquette. 'Course that'd assume you COULD provide any sort
of solution to any of the questions posed in this ng. That's unproven.

.

Loading