[GRASSLIST:1193] Re: SDTS support on grass5

[Eric G . Miller]

On Sun, Nov 26, 2000 at 12:10:04PM -0800, Ivan Kautter wrote:
> List members,
> 
> I have heard that grass5 supports the SDTS format yet I see little evidence 
> of that on the maunal pages when it comes to raster formats data such as 
> DEM.  While there seem to be commands for reading USGS data formats which 
> may be compliant with the SDTS format, the conversion process to the GRASS 
> compatible DEM is less than transparent.  I have seen various examples on 
> how to convert SDTS to GRASS including SDTS utilies offered by the BLM which 
> would not compile (there has been little concrete assistance offered by this 
> listserv or any other on this issue) and the often quoted conversion process 
> for 240K 7.5 min UGSS DEMs with 30 meter resolution involving m.read.USGS 
> and conversion to ascii format and 90 degree rotations etc.  to GRASS (which 
> I can no longer seem to locate) but not for 100K 7.5 min USGS DEMs with 10 
> meter resolution as well as m.dem.extract which may or may not be present on 
> GRASS 4.3 which I am running.

Unfortunately, no one seems to have taken up handling the SDTS DEM
format.  I have used GRASS 5 with the SDTS DLG format, and it works.
The caveat is, GRASS doesn't really handle the attribute data well, so
it's still somewhat messy.

> Can anyone offer some concrete suggestions on how to deal with SDTS format 
> data?  Should I just switch to grass5 and if so, what specifically is the 
> support offered for this format on grass5?  Additionally, if grass5 is 
> configured for floating point operations, does one need a CPU that supports 
> those type of operations to run grass5.  I have a fairly standard Pentium 
> III processor.  Is that sufficient?

I can't give you any suggestions for the DEM format.  As regards FP
rasters, an x486 is sufficient.  All Pentiums have built-in Floating
Point Unit (but at least Linux, maybe BSD, provide emulation when it
doesn't exist -- probably very slow though).

-- 
Eric G. Miller <egm2 at jps.net>