[Xastir-dev] Observations

[Tom Russo]

On Sat, Mar 19, 2005 at 07:49:59PM -0500, we recorded a bogon-computron collision of the <jerryc at netlab.org> flavor, containing:
> 
> On Thu, 17 Mar 2005, Tom Russo wrote:
> 
> > > > > > Weather objects are back to being a cause for a pause after maps
> > > > > > are loaded nad before htey are displayed.
> > > > >
> > > > > Do you mean the lit-up counties?  There are more of them on the map
> > > > > right now.
> > > > Weather Alerts, I meant..........
> > > > DUH! was still thinking object from above.
> > > 
> > > Some other things could affect this, like rtree and dbfawk perhaps.
> > > Have you changed anything with respect to those two options lately?
> > 
> > Enabling rtree should be no different from disabling it where weather 
> > alerts are concerned.  When draw_shapefile_map is called to display an alert 
> > county, all rtree indexing is skipped --- we already know which shape to 
> > display, so there's no point looking it up in the spatial index.
> > 
> > If there's a noticable difference in this behavior with and without 
> > rtree, then there's a mistake in how that skipping is done.  Please let me 
> > know if that's the case.
> > 
> Well it appears to "draw" all weather alerts weather they are in the viewable
> window or not.

That sure sounds like a bug --- have you confirmed that it's actually doing that
by enabling debug level 16?  If it's really drawing them all it would be saying
lots of lines of "found_shape ="

Does this happen irrespective of whether --with-rtree is selected?  What about
with and without dbfawk?

There is an explicit check around line 1800 of map_shp.c that should prevent
objects outside of the current view (even wx alerts) from being processed.
If it's happening anyway, there's something really screwy.

-- 
Tom Russo    KM5VY     SAR502  DM64ux         http://www.swcp.com/~russo/
Tijeras, NM  QRPL#1592 K2#398  SOC#236 AHTB#1 
 "The only thing you can do easily is be wrong, and that's hardly
  worth the effort." -- Norton Juster