Getting LIKE %you%

[Chuck Pelto]

Hi Kim,

On May 25, 2007, at 4:08 PM, Kim Kohen wrote:

> g'day Chuck,
>
>> I appreciate the offer, however, in light of the fact that V- 
>> Studio likes the code and finds the record, I think coughing up a  
>> s---pot of code would be much use.
> ••••
> Are you typing the query into VStudio or getting it from the RB  
> debugger? You really need to build the SQL in RB and copy it from  
> the RB debugger to VStudio for testing. It's starting to sound a  
> bit like an error in your code which is probably why John suggested  
> posting it.

It was my code. Albeit, not so much my code as my misunderstanding of  
the idiotsyncracies of SQL.

I was not clearing the previous VCursor before doing my call to find  
records.

Coming from an environment, like FMP, it is not necessary to clear a  
search before performing a new search. Or, at least, it is done  
behind the scenes.

The clue to resolve the dilemma was the comment from yesterday where  
someone mentioned seeing the non-annotated error in another venue and  
touching on that aspect.

Doing a NoLock search on the previous search string and declaring the  
VCursor as 'nil', did the trick.

Cumbersome, but effective.

This solution, as usual, brings up another question. But I won't  
address it now. Maybe later next week.

Thanks for all the help.

Regards,

Chuck
P.S. Ruslan....

....you might want to improve the current methods for finding Error  
Codes. There seem to be some interesting issues with the XML  
approach. Furthermore, I think the 460037 codes absence is something  
of an issue. You might want to at least point to where people can  
find the language description of what the number means.