[GRASS5] Re: [Gdal-dev] GDAL CVS Instability - Persistant Auxilary Metadata
Hamish
hamish_nospam at yahoo.com
Wed Apr 27 22:20:43 EDT 2005
On Wed, 27 Apr 2005 12:23:33 -0400
Frank Warmerdam <fwarmerdam at gmail.com> wrote:
> Folks,
>
> First, this is a warning that GDAL CVS will likely be somewhat
> unstable for a couple weeks. I am starting to commit substantial
> changes to provide Persistant Auxilary Metadata to many formats,
> including read-only formats.
>
> Persistent Auxilary Metadata is basically intended as a way of
> associating an extra header file with files that can store various
> information that isn't supported by the format itself. That is
> auxilary metadata like coordinate systems, geotransforms, histograms,
> generic metadata, color tables, and so forth.
>
> The auxilary metadata is stored in a .pam XML file. The .pam filename
> is the original dataset name plus .pam. So the pam file for utm.jpg
> would be utm.jpg.pam.
>
> The general idea is to ensure we can associate a variety of auxilary
> information (and cache statistics) for most GDAL supported formats,
> whether they have support for such structures in the format itself or
> not.
>
> The XML structure is similar to how such information was stored in
> .vrt files, for those familiar with vrt.
>
> The process to PAM enable a driver is to derive it from GDALPamDataset
> and GDALPamRasterBand instead of GDALDataset and GDALRasterBand
> and then to make a few minor adjustments in the driver. It is my
> intention to apply this change to almost all the drivers.
>
> In the meantime, if you update from CVS please also do a clean
> rebuild.
>
> I plan to write up a tutorial document describing how PAM works
> fairly soon.
I think GRASS maps could benefit from this as well. Good|some|any
meta-data support has long been missing in GRASS and this might be
the answer. Synergy & other buzz words plus. Looking forward to
further information on the implementation.
Hamish
More information about the grass-dev
mailing list