On Fri, Sep 12, 2008 at 04:50:00PM +0200, Andreas Kalsch wrote:
> Subject: Re: [OSM-dev] Spatial vs. multi-column indexes for points
> Thanks!
>
> And this really outperforms a simple query like "... latitude between
> y1 and y2 and longitude between x1 and x2" ?
> Florian Lohoff wrote:
> > On Fri, Sep 12, 2008 at 08:20:18AM +0200, Andreas Kalsch wrote:
> >
> >> Subject: Re: [OSM-dev] Spatial vs. multi-column indexes for points
> >>
> >>
> >>> On Thu, Sep 11, 2008 at 08:31:23PM +0200, Andreas
Florian Lohoff wrote:
On Fri, Sep 12, 2008 at 08:20:18AM +0200, Andreas Kalsch wrote:
Subject: Re: [OSM-dev] Spatial vs. multi-column indexes for points
On Thu, Sep 11, 2008 at 08:31:23PM +0200, Andreas Kalsch wrote:
All, thanks for your quick responses!
Quad tiles look like
On Fri, Sep 12, 2008 at 08:20:18AM +0200, Andreas Kalsch wrote:
> Subject: Re: [OSM-dev] Spatial vs. multi-column indexes for points
>
> > On Thu, Sep 11, 2008 at 08:31:23PM +0200, Andreas Kalsch wrote:
> > > All, thanks for your quick responses!
> > >
> > &
> > My current optimization includes:
>
> >
> > - using mediumint for lat/lon - enough for ~2 meters resolution
> > - using a bounding box first for point+radius calculation and then
> selecting the circle with a Pythagoras approximation, which is exact enough
> >
> > For the first, I want to us
> On Thu, Sep 11, 2008 at 08:31:23PM +0200, Andreas Kalsch wrote:
> > All, thanks for your quick responses!
> >
> > Quad tiles look like a smart way to create an index. So to lookup a
> > single point or a quad tile, this is fine. But for my application I
> > need another lookup - by bounding box
> My current optimization includes:
>
> - using mediumint for lat/lon - enough for ~2 meters resolution
> - using a bounding box first for point+radius calculation and then selecting
> the circle with a Pythagoras approximation, which is exact enough
>
> For the first, I want to use MySQL.
>
>
On Thu, Sep 11, 2008 at 08:31:23PM +0200, Andreas Kalsch wrote:
> All, thanks for your quick responses!
>
> Quad tiles look like a smart way to create an index. So to lookup a
> single point or a quad tile, this is fine. But for my application I
> need another lookup - by bounding box with any rat
On Thu, Sep 11, 2008 at 1:31 PM, Andreas Kalsch <[EMAIL PROTECTED]>wrote:
> All, thanks for your quick responses!
>
> Quad tiles look like a smart way to create an index. So to lookup a single
> point or a quad tile, this is fine. But for my application I need another
> lookup - by bounding box wi
All, thanks for your quick responses!
Quad tiles look like a smart way to create an index. So to lookup a single
point or a quad tile, this is fine. But for my application I need another
lookup - by bounding box with any ratio and size. Is there a way to look up a
special bounding box with this
On Thu, Sep 11, 2008 at 01:49:55PM +0200, Andreas Kalsch wrote:
> Subject: [OSM-dev] Spatial vs. multi-column indexes for points
>
> Hey,
>
> last week I made some experiments with huge datasets of lat/lon
> points. I use MySQL 5.0, which partially support GIS extensions,
> including R-trees. But
On Thu, Sep 11, 2008 at 7:23 AM, Marcus Wolschon <[EMAIL PROTECTED]>wrote:
>
> Hello.
>
> Is there amy kind of embeddable (java) database or
> database I can expect an end-user to install (no PostGIS or manually
> installing
> the mysql-gis-extension) that supports spatial indice?
>
I've looked f
Hello.
Is there amy kind of embeddable (java) database or
database I can expect an end-user to install (no PostGIS or manually installing
the mysql-gis-extension) that supports spatial indice?
I tried storing a z-tree of the coordinates in one indexed column
in HSQLDB and the performance was wor
On Thu, Sep 11, 2008 at 01:49:55PM +0200, Andreas Kalsch wrote:
Has anybody used GIS successfully in MySQL or PGSQL and can tell me
how the performance compares between the two techniques?
At least in PostgreSQL/PostGIS, the geometric indices are WAY faster.
Just remember to either use the ST_*
Instead of indexing by precise coordinates you could index by virtual
tiles as it is done in OSM's main DB since a year ago with nice
performance boost:
http://wiki.openstreetmap.org/index.php/QuadTiles
good luck,
Stefan
On Thu, Sep 11, 2008 at 1:49 PM, Andreas Kalsch <[EMAIL PROTECTED]> wrote:
15 matches
Mail list logo
