Re: Painfully slow searching

[Joe Pluta <joepluta@xxxxxxxxxxxxxxxxx>]

Brian Johnson wrote:
> Even considering network overhead of a client-server app, it's hard to see
> how this searching feature could have been implemented in a way to be any
> SLOWER, especially when searching a list of members.  I have gone to the
> step of downloading all source into an iSeries project just for searching.
>   

I really don't want to be an apologist for this, because I agree that 
the search facility is slower than I would like.  As you've mentioned, 
though, Brian, there are alternate methods depending on what it is 
you're trying to do.

I guess my question is how often are you doing these searches, and why?  
I do indeed occasionally do a text search, but it's rare. Are you doing 
this once a month?  Once a week?  Once a day?  Once an hour?  It seems 
that unless this is something you're doing over and over the few minutes 
spent aren't going to impact your daily job.

On the other hand, if you are in the situation where you have to use 
string scans repetitively, then why are you doing it?  Are you doing 
impact analysis?  Looking for file/field where used information?  
Internationalization?  Searching for mod markings (used to do that all 
the time back in a previous life)?

If so, is it possible that you need a different tool to help you with 
that?  Something that performs a more in-depth analysis and gives you 
better information?  I'm just asking, because sometimes bumps in the 
road surface because of other issues.  We always use to say in the ERP 
game that lowering the inventory levels found the problems in the 
process.  Maybe this is an opportunity to improve a process?

Or maybe it's just that the search is too slow, and we need to ask IBM 
to beef it up a bit.

Joe