Index DESCEND with clipper 5..3b 
Author Message
 Index DESCEND with clipper 5..3b

Quote:

> I'm having a lot of problem with one index using
> DESCEND(STR(Profit,14,2)).

Clipper 5.2x+ has a descend clause on the INDEX expression, also recognized by
Advantage.

index on str( profit,14,2) descending to dfgdgf

or some such...



Sun, 04 Jun 2000 03:00:00 GMT  
 Index DESCEND with clipper 5..3b

Quote:

>Hello all,

>I'm using clipper 5.3b with Blinker 4.10 on a Novell 4.11 env. Advantage
>Database server 4.10 is also used in my env.

>I'm having a lot of problem with one index using
>DESCEND(STR(Profit,14,2)). It giving me random GPF on the update of this
>index. The problem seems to be in a Clipper function. What can I do to
>solve this problem?

>I really need a descending index...

>Here bellow, the place where it crashes:

>$

<snip>

Clipper's Descend() function has been broken for
some time ( first reported 11/94). A replacement,
FT_Descend, is available - see Descnd.zip on
CIS and/or maybe on the Oasis? If you have no
luck, contact me via email and I'll send you a
copy.

Jo



Mon, 05 Jun 2000 03:00:00 GMT  
 Index DESCEND with clipper 5..3b

Quote:

> <snip>

> Clipper's Descend() function has been broken for some time
> ( first reported 11/94). A replacement, FT_Descend, is available
> - see Descnd.zip on CIS and/or maybe on the Oasis? If you have no
> luck, contact me via email and I'll send you a copy.

> Jo

Gee, that's odd. I've been using DESCEND() in several indexes for the
past 7 years,
never had a problem with them. Some of these indexes are on a DBF of
500,000 recs,
the NTX is over 10meg in size. Using 5.01 & Blinker 3.x on Netware 3.11,
I have
NEVER had a corrupted index. Never. I'm going to miss NTX'es when I
finish moving
over to windows programming.

  Ron Cherry

Remove KILLSPAM. from address to return email to me (not necessary in
NG's).



Mon, 05 Jun 2000 03:00:00 GMT  
 Index DESCEND with clipper 5..3b

On Thu, 18 Dec 1997 22:22:17 -0800, Ron Cherry

Quote:


>> <snip>

>> Clipper's Descend() function has been broken for some time
>> ( first reported 11/94). A replacement, FT_Descend, is available
>> - see Descnd.zip on CIS and/or maybe on the Oasis? If you have no
>> luck, contact me via email and I'll send you a copy.

>> Jo

>Gee, that's odd. I've been using DESCEND() in several indexes for the
>past 7 years, never had a problem with them. Some of these indexes
>are on a DBF of 500,000 recs, the NTX is over 10meg in size.
>Using 5.01 & Blinker 3.x on Netware 3.11 I have NEVER had a
>corrupted index. Never....

Which just goes to show that your decision to stick with
5.01 was in your favor in this instance.:-)


Tue, 06 Jun 2000 03:00:00 GMT  
 Index DESCEND with clipper 5..3b

Quote:


> >> <snip>

> Which just goes to show that your decision to stick with
> 5.01 was in your favor in this instance.:-)

I've thought so for about 5 years.

  Ron Cherry

Remove KILLSPAM. from address to return email to me (not necessary in
NG's).



Tue, 06 Jun 2000 03:00:00 GMT  
 Index DESCEND with clipper 5..3b

Quote:
> > Clipper's Descend() function has been broken for some time
> > ( first reported 11/94). A replacement, FT_Descend, is available
> > - see Descnd.zip on CIS and/or maybe on the Oasis? If you have no
> > luck, contact me via email and I'll send you a copy.
> Gee, that's odd. I've been using DESCEND() in several indexes for the
> past 7 years, never had a problem with them.

Jo will have to post the test code that shows how it's broke, if he
still has it. It's simple, and it shows it's broke. The FT_DESCEND()
function from NANFOR.LIB was written in response to it's being broken.
The Clipper DESCEND() function won't work with the Comix RDD because of
this defect whilst the FT_DESCEND() function works perfectly.


Tue, 06 Jun 2000 03:00:00 GMT  
 Index DESCEND with clipper 5..3b

Quote:


>> I'm having a lot of problem with one index using
>> DESCEND(STR(Profit,14,2)).

>Clipper 5.2x+ has a descend clause on the INDEX expression, also recognized by
>Advantage.

>index on str( profit,14,2) descending to dfgdgf

>or some such...

If you suspect that the DESCEND() is causing the problem,
then try doing a manual descend to prove it.

* using an offset (10000000) value that is higher than the
* largest number in the field causes a descending order

   index on str(1000000-profit)+name to whatever

Give it a try.

See ya,

Steve



Fri, 09 Jun 2000 03:00:00 GMT  
 Index DESCEND with clipper 5..3b

Quote:

>> I'm having a lot of problem with one index using
>> DESCEND(STR(Profit,14,2)).

>Clipper 5.2x+ has a descend clause on the INDEX expression, also recognized by
>Advantage.

>index on str( profit,14,2) descending to dfgdgf

>or some such...

I have had problems using descend().  Usually with date fields.  

Original index string under S87 and 5.2e  DESCEND(DTOS(date))

Under 5.3 this would occassionally GPF.  Same location in dbf when
reindexing.  

Changed index expression to STR(DESCEND(date),7).
(Descend on a date field yields a 7 digit number. )

This has solved my problem.  

Steven Meyer
DEK Software, Inc.
Hoffman Estates, IL



Mon, 26 Jun 2000 03:00:00 GMT  
 
 [ 12 post ] 

 Relevant Pages 

1. Index DESCEND with clipper

2. Using Foxplus indexes with clipper 5.3b

3. Clipper...Clipper...Clipper - 5.3b Patch

4. Search with Descend Index

5. index order descending in one argument only

6. Descending problems in index COMIX driver

7. error DESCEND() in CDX Index file

8. Descend(field) of CDX index

9. Using DESCENDING clause with Index command

10. Descending index

11. Descending index with dbcreateindex()

12. Clipper 5.01 and Clipper 5.3b NTX compatibility

 

 
Powered by phpBB® Forum Software