[gdal-dev] Geotiff FillEmptyTiles with no data value?

[Even Rouault]

Le mercredi 06 juillet 2016 16:17:32, Daniele Romagnoli a écrit :
> Hi Even,
> thanks I'm already using the SPARSE_OK=TRUE to create an empty tiff
> and afterwards warp input files on that with SKIP_NOSOURCE=YES.
> I have already updated the ImageIO-EXT's TIFFImageReader to deal with
> tile/strip with zero byte count.
> 
> Final question:
> Starting from which GDAL version, SPARSE_OK=TRUE and SKIP_NOSOURCE=YES are
> supported?

Woo, a long time. I wasn't probably even born when they were added. 

Digging a bit in history shows that SPARSE_OK was added in 2008 / GDAL 1.6 and 
SKIP_NOSOURCE in 2006...
Returning the nodata value for implicit blocks instead of 0 was added in 1.7.

> 
> Thanks for your feedbacks.
> Cheers,
> Daniele
> 
> 
> On Wed, Jul 6, 2016 at 3:53 PM, Even Rouault <even.rouault at spatialys.com>
> 
> wrote:
> > Daniele,
> > 
> > > has this improvement also being backported on previous versions (as an
> > > instance 2.1.x, 2.0.x)?
> > 
> > No, trunk only.
> > 
> > You could as well use the SPARSE_OK=TRUE creation option to avoid
> > writting blocks at all. Provided that your reader knows how to deal with
> > tile/strips with a 0 byte count (GDAL based readers back to the older
> > days are OK)
> > 
> > Even
> > 
> > > Cheers,
> > > Daniele
> > > 
> > > On Tue, May 10, 2016 at 3:29 PM, Even Rouault <
> > 
> > even.rouault at spatialys.com>
> > 
> > > wrote:
> > > > Le mardi 10 mai 2016 09:22:19, jramm a écrit :
> > > > > Hi
> > > > > I have made changes to FillEmptyTiles so that if nodata is set,
> > > > > then
> > 
> > it
> > 
> > > > > will always fill with nodata, otherwise 0.
> > > > > 
> > > > > I have attached the raw diff...I have no idea how to submit a
> > > > > change request/review etc?
> > > > > 
> > > > > fillempty_nodata.diff
> > > > > <
> > 
> > http://osgeo-org.1560.x6.nabble.com/file/n5265514/fillempty_nodata.dif
> > 
> > > > > f
> > > > 
> > > > I've committed in trunk a more complete version of your patch since
> > > > on reflection there was also the case of non power-of-two bit depths
> > > > to handle.
> > > > 
> > > > Patches are better handled as attachments to Trac tickets.
> > > > 
> > > > Even
> > > > 
> > > > --
> > > > Spatialys - Geospatial professional services
> > > > http://www.spatialys.com
> > > > _______________________________________________
> > > > gdal-dev mailing list
> > > > gdal-dev at lists.osgeo.org
> > > > http://lists.osgeo.org/mailman/listinfo/gdal-dev
> > 
> > --
> > Spatialys - Geospatial professional services
> > http://www.spatialys.com

-- 
Spatialys - Geospatial professional services
http://www.spatialys.com