Really slow processing

[Jeff Harrison]

----- Original Message -----

> From: Jeff Harrison <jeffaharrison at yahoo.com>
> To: Bob <roberth at sim-soft.com>; fplist <filepro-list at lists.celestial.com>
> Cc: 
> Sent: Friday, August 26, 2011 1:29 PM
> Subject: Re: Really slow processing
> 
> ----- Original Message -----
> 
>>  From: Bob <roberth at sim-soft.com>
>>  To: fplist <filepro-list at lists.celestial.com>
>>  Cc: 
>>  Sent: Friday, August 26, 2011 12:57 PM
>>  Subject: Really slow processing
>> 
[snip]

>>  Any suggestion on speeding the process?
>> 
>>  Oh, they are running fp 5.0.9 I believe.
>> 
>> 
>>  Bob
>>  "The nonGeek Geek"
> 
> This is a good example of what not to do on a few fronts.  First of all, the 
> inkey is almost certainly slowing you down, and It has no purpose that I can see 
> (other than slowing you down).  Secondly, good, you got rid of the close - that 
> is a start.  It does not make sense to keep looking up the same record, then 
> close it - for each record that is read.  setvar only needs to run once - or not 
> at all if you make rn global and move the "rn(1,.0,g)=1" to the 
> "enter" subroutine.  Also, you can then move the getmst into the 
> "enter" routine and make the xd and xe fields global as well if they 
> are not already.
> 
> Also, I see that you are writing to the master file - clearing out fields 14 and 
> 15 - is that intended?  If f so, then you should protect the lookup and you 
> should execute a write master (or uncomment the close).  This would not be a 
> speed issue as you would only be executing this once  - instread of  the 1 
> million times that you are currently getting.
> 
> That is all that I see from a quick look - but thats a bunch!
> 

Another thing - You should probably look into using a lookup dash so that you can utilize the indexes and do a lightning fast selection.  I'm sure that there are faq's on this technique somewhere.

Jeff Harrison
jeffaharrison at yahoo.com
Author of JHExport and JHImport