[OpenLayers-Dev] singleTile optimization

[Erik Uzureau]

There we go. What David's said here is what I was grasping for earlier here.

If you use a singleTile layer that is gridded, *eventully* the user is
going to pan the map such that the corners of the grid are in the
viewbox. At that point, the only option is to load 4 tiles, at which
point, we're back to gridded, in which case you might as well just use
a normal gridded layer with a very big tilesize and a buffer:0

Alternatively, One could imagine a tile layer which is double-tiled
and that would avoid this problem, but seems like a horrible
implementation nightmare.



On 10/4/07, David William Bitner <david.bitner at gmail.com> wrote:
> Ducky!
>
> Great to see you here -- don't be a stranger.
>
> The problem with aligning singleTile to a grid is that as soon as you have
> the corner of the grid on the screen, all the sudden you are in a tiled mode
> only with larger tiles.  It strikes me that for any use case that would
> require any kind of grid alignment at all, really you should be using a
> tiled mode only with perhaps larger tiles and smaller buffers.  The use case
> that I see for single tile is for dynamic data -- data that is dynamic
> enough that you wouldn't want it cached at all, ever.
>
> David
>
>
> On 10/4/07, Dan Little <danlittle at yahoo.com> wrote:
> >
> >
> >
> > Yikes, first time poster!
> >
> > I think the speed to generate a single tile is wholely dependent on the
> map-rendering engine and the dataset.
> >
> > For example, here at the City of Saint Paul, our streets layer is entirely
> vector.  I can generate a very very large tile with highly detailed streets
> for the entire city using Mapserver in less than 1/3 of a second.  Given the
> nature of MapServer, and the size of the shapefile which stores our streets
> layer, multiple tiles is much much slower (as the shapefile needs to be read
> from disk and loaded to memory for each rendered tile).  It would be nice to
> have some of the effects of tiling without the high-cost it currently
> presents in rendering performance.
> >
> > Dan "Ducky" Little
> > City of Saint Paul / Public Works / GIS
> >
> >
> >
> >
> > ----- Original Message ----
> > From: Stephen Woodbridge <woodbri at swoodbridge.com>
> > To: pspencer at dmsolutions.ca
> > Cc: dev at openlayers.org
> > Sent: Wednesday, October 3, 2007 9:37:29 PM
> > Subject: Re: [OpenLayers-Dev] singleTile optimization
> >
> >
> > pspencer at dmsolutions.ca wrote:
> > > What do you think about aligning singleTile bounds to the grid to
> > > promote cacheability in the client?
> >
> > I think that this should be an option in the layer definition if it gets
> > done. I can see some benefits to doing this, but the down side is that
> > all images generated will be larger and slower to generate and the
> > singleTile option needs to support more than one use case in my mind
> anyway.
> >
> > -Steve
> > _______________________________________________
> > Dev mailing list
> > Dev at openlayers.org
> > http://openlayers.org/mailman/listinfo/dev
> >
> >
> > ________________________________
> Luggage? GPS? Comic books?
> > Check out fitting gifts for grads at Yahoo! Search.
> > _______________________________________________
> > Dev mailing list
> > Dev at openlayers.org
> > http://openlayers.org/mailman/listinfo/dev
> >
> >
>
>
>
> --
> ************************************
> David William Bitner
> _______________________________________________
> Dev mailing list
> Dev at openlayers.org
> http://openlayers.org/mailman/listinfo/dev
>
>