[GRASS5] A naive opinion on how grass *should* work

[Glynn Clements]

Russell Nelson wrote:

>  > set up a region at the beginning is a surprisingly big obstacle.
>  > (well not so surprisingly, given the oldish interface for doing that)
> 
> No, it's not the interface.  Well, okay, it *is* the interface, but
> more than that it's the necessary to know MORE than I really know.
> Here's the questions I asked myself:
> 
> ``1??  They're giving me a default lat/lon of 0N, 0E through 1N, 1E??

Not quite; it's the origin of the current coordinate system, which may
or may not be a lat/lon coordinate system.

The image has to go somewhere. The alternative is for all of the
programs which import uncorrelated images to insist that the user
specifies the boundaries on the command line.

> ``Cell resolution?  What's a cell?  How do I know what my cell
> resolution is?  I haven't even imported my first map, and it wants to
> know what the resolution of the map is going to be?  I have no
> fricking idea.''

The user may or may not know what resolution they want, but GRASS
absolutely cannot know. The same applies to the region boundaries.

Note that the resolution specified when creating a location is the
default for the resolution setting which GRASS uses when creating new
raster maps. This doesn't have to correspond to the resolution of any
maps which you import. And you can change the actual output resolution
from its default value with g.region.

For vector and sites maps, the resolution doesn't matter.

> ``Hrm.  It doesn't accept "75.00W".  Maybe "75 00'W".  No, that's not
> right either.  How about "-75.00"?  No, that doesn't work either.  How 
> about "75.0000W"?  How about "7500W"?  ARRGGGGGHHHHHH!''
> 
> ``Ahhhh, okay, it just wants integer degrees.  But how big should I
> make my region?  Just the one map I want to do first, or should I make 
> it big enough to hold all fifty maps??''

Again, you can change the actual region boundaries at any time with
g.region.

>  > I guess that the problem to getting this addressed is that most of the
>  > issues mentioned below are not such big issues once you use GRAS a lot
>  > and especially with the improvements in r.in.gdal. (Russ use r.in.gdal
>  > instead of r.in.tiff).
> 
> It didn't get built on my machine because libgdal was missing.

By default, r.in.gdal should be built regardless of whether you have
libgdal. However, it won't actually run unless libgdal is present.

>  > For example the issue of adapting the region
>  > to displayed map (as Carl has mentioned before) would drive you crazy
>  > if you are doing some modeling using a larger number of maps with
>  > different extent and resolutions and you need to preview those maps,
> 
> Wait a second.  You mean that the region is only used for viewing??

No. As well as viewing, the region is also used when generating new
raster maps (which is probably *the* most important function of
GRASS).

-- 
Glynn Clements <glynn.clements at virgin.net>