[GRASS5] Re: LZW removal

[Markus Neteler]

Hi developers,

please assist Eric to test the new "flate" compression!
I am using it at home without problems. Certainly you
should make a backup of your data.

How to "upgrade":

- Uncompress all data in "old GRASS 5":
  r.uncompress -u rastermap

- compile GRASS with flate method
  cd src/libes/gis/
  edit get_row.c put_row.c
  comment in both files: "#define USE_LZW_COMPRESSION"

- run make install to affect the lib and all modules

- Try it: You can compress the uncompressed data then with
  r.uncompress rastermap


We want to activate flate compression as default on 
Dec. 17th, so a few days to report eventual errors are left.

Thanks

 Markus



------ fwd -------
On Sun, Dec 10, 2000 at 10:00:56AM -0800, Eric G . Miller wrote:
> On Sun, Dec 10, 2000 at 05:02:56PM +0000, Markus Neteler wrote:
> > > But the whole idea is to remove all of the LZW code from GRASS.  My
> > > understanding is this will only affect people who have used earlier
> > > versions of GRASS 5, so for users upgrading from GRASS 4, this won't be
> > > an issue.  I know it's a bit confusing and annoying for the user, but I
> > > don't know how to simultaneously remove LZW from the GRASS codebase and
> > > include it ;).
> > Oh! You mean that LZW was introduced in GRASS 5? I didn't know that.
> > So the GRASS 4.x didn't have LZW?
> 
> I don't know exactly when it was introduced, there's no date or author
> on the lzw.c file.  However, CELL file use a totally different encoding
> and compression scheme (RLE based).  Only FP rasters use XDR encoding
> and LZW compression.  G3D uses an encoding similar to CELL files, with
> optional RLE and LZW compression methods (I don't quite understand how
> it works).
> 
> > > My recommendation is that people run r.compress -u on all their FP
> > > rasters before upgrading.  I think it is the surest path. Exporting to
> > > ASCII should not be necessary.
> > I see.
> 
> Yes, you can still preserve all the color and cats info...
> 
> > > > I'll keep working with flate now (currently I'm updating my satellite image
> > > > radiometric/terrain correction  module to GRASS 5, big job for me...).
> > > > Maybe several problems interfere from time to time :-)
> > Up to now I didn't face any problems. Perhaps these stripes have been in the
> > image. Currently I don't want to recompile again to LZW...
> 
> Well, let me know.  I wish we could get more people to try this out.
> 
> -- 
> Eric G. Miller <egm2 at jps.net>

---------------------------------------- 
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo at geog.uni-hannover.de with
subject 'unsubscribe grass5'