Can anyone tell me what the differenceis between uing EQ and using IS in a FEX?

Thanks in advance!

This message has been edited. Last edited by: Kerry,

none

Hi,

How can it be? There should be a reason!!!

Regards.

There are a lot of mysteries in WebFOCUS code.

Using Operators in Record Selection Tests

What about this little tidbit: "WRITE and ADD are synonyms of SUM and can be substituted for it"?

Using Display Commands in a Request

Another such 'mystery' is that you can add an 'S' to any fieldname in your table request and the code will still run fine.
TABLE FILE CAR PRINT CAR is the same as TABLE FILE CAR PRINT CARS ...

This one I didn't know!

I have never appreciated being able to do this:

TABLE FILE CAR
PRINT
F L A RE
END

Who remembers LET?

LET TFC = TABLE FILE CAR PRINT F L A RE
TFC
END

T

I recently posted a smart@$$ answer to a LET question:

LET KINKS = CAR
LET STONES = FILE
LET ANIMALS = *
LET BEATLES = TABLE
LET HOLLIES = END
LET SHADOWS = PRINT

BEATLES STONES KINKS
SHADOWS ANIMALS
HOLLIES

You're showing your age now Francis

T

I know, I shouldn't have mentioned The Shadows!

You know there has been some good music since the 60's

Yes, I love TV On The Radio.

i'm now tempted to write a function for my edasprof, next to my LETs , called
GOBBLEUP
and another one called SPITITOUT
wait that's more than 8 characters
how about
SPITITUP
blech
just to train the truly gullible ....
today this consultant who just got the axe from IB actually asked me how to concatenate strings , and as if that weren't bad enough, What manual, and Where could he find one?

Susannah,

That's really sad. I used to work for IB and I can assure you I know where all the manuals are and I read them regularly.

quote:

TABLE FILE CAR
PRINT
F L A RE
END

For those who don't know what Francis was doing here, he was using what is called the 'shortest unique truncation'. This is great for minimalists and those that don't like to type.

quote:

shortest unique truncation

Thanks for giving it a name!

Francis isn't showing his age by mentioning all those groups. He's showing good musical taste and the wisdom to know great music from mediocre music.

i second that!

To get back to the original topic - the difference used to be in what SQL was generated - in my experience anyway!

Consider the two pieces of code -

SUM CNT.fieldname

WRITE CNT.fieldname

Now the more experienced among us would say that they should produce the same results, and I would agree. But the devil is in the detail and it is how the results are retrieved that differed.

The first one (certainly used to against DB2 on OS390) extracted all the records from the SQL source and then let FOCUS do the counting. Double handling of the data and costly in I/O terms. The SQL that was produced (circa 1994) was SELECT * FROM ..... and the count was achieved in FOCUS.

The second one (same as previous comment) would produce the SQL SELECT COUNT(*) FROM ..... - getting the SQL to do all the counting work, single handling of data etc. etc. etc.

So my answer would be that if you are reporting against an SQL source then trap the SQL and see if there is a difference. Chances are that since 1994 (cough) the back end would most certainly have improved the interpretation of FOCUS to SQL and maybe now there is no difference other than keystrokes, however we all know that to have truely backward compatibility legacy syntax must be retained.

How close I am to the view of IB only the Product Managers will be able to confirm (I know that at least two of them are from that time ).

T

For that matter, ADD is (still) a valid synonym of WRITE:

TABLE FILE CAR
ADD SEATS LENGTHS HEIGHTS
BY COUNTRY
END

(Where's my TTY?)

DB2 v9 on AIX, WF v7.6.5.

Both SUM and WRITE produce the same SQL:

SELECT T1.TIME_DIM_KEY, COUNT(T1.PERIOD_END_DT)
FROM TIMED T1
GROUP BY T1.TIME_DIM_KEY
ORDER BY T1.TIME_DIM_KEY
FOR FETCH ONLY;

Francis,

I guessed that it would which is why I mentioned DB2 on OS390 circa 1994!

T

Yes, I've gathered that they've cleaned things up since 1994