>Using the Upper() in the index is sometimes problematic because it makes Rushmore optimization a little trickier. (You have to always use the upper() in the operation.) Nonetheless, it is probably a good idea to use it with a client name. Anyway, Quinn - here's the solution all put together and guaranteed to work:
>
>Now:
>1. Select Client && Assuming client is the name of your client table
>2. Index on Upper(Cl_Name) Tag UCl_name && cl_name is the client name field
>
>
>Later, in your code:
>1. Make sure the client table is in use in Some workarea.
>2. l_lRetVal = ;
> Seek(PadR(Upper(txtTextBox.Value),FSize('cl_name','clients'));
> ,'clients','UCl_name')
>
>If that doesn't work, shoot me.
If this doesn't save your life, don't shoot me: using other collating sequences (like "General", "uniqwt" or any national sequence) creates indexes where upper and lower are not distinguished, so this wouldn't have happened in the first place. This way it really doesn't matter if you seek(lower(locclient)), seek(proper(locclient)) or seek(upper(locclient)) - it finds it all three times if it exists.
