FW: Browse lookup - odd behavior
Jay R. Ashworth
jra at baylink.com
Fri Jul 16 09:53:40 PDT 2004
On Fri, Jul 16, 2004 at 12:47:08PM -0400, Kenneth Brody wrote:
> "Jay R. Ashworth" wrote:
> [...]
> > > Or are you suggesting actually re-running the lookup in its entirety?
> >
> > Doing anything else seems semantically inadequate, as this problem
> > demonstrates.
>
> After some investigation last night, it appears that I misrembered how
> the pkeep part of the browse lookup works.
>
> Upon re-executing the pkeep lookup, it is only the initial record whose
> position is cached. From there, the rest of the browse window is repopulated
> by scanning the file.
Hmmm... as fast as index lookups have always been (and I'll grant ya'll
this: index lookups have *always* been breathtakingly fast, even on
Xenix, on a 386-16 :-), did you actually win much by doing that?
> While within the browse, the record numbers are cached so that scrolling
> within the window does not re-scan the file. Also, when scrolling off the
> end of the window, everything is shifted within the cache, and only the
> newly-vacated area is populated by continuing the scan of the flie.
Ah.
> Since my misremembering of the caching fit the symptoms being reported, I
> thought that I remembered correctly.
Happens.
> Which brings us back to the OP's symptoms.
>
> We would need more specifics of what the index is built on, which record
> is being deleted, what is in the newly-created record that's replaced it,
> specifically the field(s) used to build that index, what records appear in
> the window upon re-executing the browse lookup, and perhaps some more
> information that I'm not thinking of at this moment.
Which suggests that there *is* something broken, maybe.
Cheers,
-- jra
--
Jay R. Ashworth jra at baylink.com
Designer Baylink RFC 2100
Ashworth & Associates The Things I Think '87 e24
St Petersburg FL USA http://baylink.pitas.com +1 727 647 1274
"You know: I'm a fan of photosynthesis as much as the next guy,
but if God merely wanted us to smell the flowers, he wouldn't
have invented a 3GHz microprocessor and a 3D graphics board."
-- Luke Girardi
More information about the Filepro-list
mailing list