[postgis-devel] [PostGIS] #2122: [raster] Real extent feature lost after metadata as views

[Sandro Santilli]

On Mon, Dec 03, 2012 at 04:58:47PM +0000, Mateusz Loskot wrote:
> On 3 December 2012 16:36, Sandro Santilli <strk at keybit.net> wrote:
> > On Mon, Dec 03, 2012 at 04:29:14PM +0000, Mateusz Loskot wrote:
> >> On 3 December 2012 16:05, Sandro Santilli <strk at keybit.net> wrote:
> >> > On Mon, Dec 03, 2012 at 03:13:00PM +0000, Mateusz Loskot wrote:
> >> >> On 1 December 2012 16:03, Sandro Santilli <strk at keybit.net> wrote:
> >> >>
> >> >> Despite technically possible tohandle, it would complicate things though.
> >> >> Is actual blocksize < max_blocksize_x|y allowed only along the right and
> >> >> the bottom edge?
> >> >
> >> > No, I think it should be allowed everywhere
> >>
> >> Then, we no longer discuss blocking in a regular grid.
> >
> > It'd be a regular-constrained grid
> 
> I don't understand what you mean.
> Constraint how?

By the max width and max height of each tile.

> >> > I guess GDAL could just ensure the same blocksize when reading the rasters.
> >> > There's an ST_Resize function IIRC.
> >>
> >> This is optimisation feature, cache feature, predictable memory constraint, etc.
> >
> > The max size constraint would still give you predictable memory.
> > You should be more specific about the cache and optimisation parts.
> 
> I speak of GDAL metchanisms and notion of natural block.

I think that's to reduce the number of I/O calls, and I don't see
how existance of smaller tiles would break that. Note that I'm still
envisioning that each tile in such a "regular tile thing" would have
its upper and leftmost edge anchored to the regular grid.

--strk;